National and International Agricultural Genome-to-Phenome

JENNIFER CLARKE Author

03/03/2023 Added

22 Plays

Description

Professor of Statistics and Food Science and Technology; Director, Quantitative Life Sciences Initiative, University of Nebraska–Lincoln

Searchable Transcript

Search:

[00:00:00.780]The following presentation is part
[00:00:02.730]of the Agronomy and Horticulture Seminar Series
[00:00:05.820]at the University of Nebraska-Lincoln.
[00:00:08.160]So thank you everybody
[00:00:08.993]for coming this afternoon.
[00:00:11.940]I'm Jessica Peterson.
[00:00:12.773]I'm one of the faculty members in Breeding and Genetics
[00:00:14.910]over in the Animal Science Department.
[00:00:17.400]So I'll give you a little background
[00:00:18.510]of why we're all here in the same room.
[00:00:22.140]Historically, animal science and agronomy
[00:00:25.010]in the genetics world have worked together
[00:00:27.720]teaching some classes, joint teaching some graduate classes,
[00:00:31.170]and in talking about ideas together.
[00:00:33.930]And more recently, I was on a search committee,
[00:00:35.970]we've hired in progress of hiring a new animal science
[00:00:39.990]and agronomy quantitative geneticists.
[00:00:43.770]And in those conversations
[00:00:45.630]and thinking about our curriculum,
[00:00:47.250]we were thinking about,
[00:00:48.270]boy, there's a lot of ways we overlap,
[00:00:50.370]not just in genetics, but in phenotyping,
[00:00:52.980]in climate questions and interacting with stakeholders
[00:00:56.700]and all of those ideas.
[00:00:58.620]And it was actually during one
[00:01:01.140]of those candidate interviews where he asked us,
[00:01:04.597]"Do you have a joint seminar?"
[00:01:06.780]And we thought we don't, but maybe we should
[00:01:10.200]because we do share a lot of the same interests
[00:01:12.420]and there are opportunities where we can get money
[00:01:16.320]to share these interests and do research together.
[00:01:20.460]And so when we're thinking about a person to speak
[00:01:22.800]for our first joint seminar,
[00:01:25.470]Dr. Jennifer Clarke came to mind
[00:01:27.180]because she works in this space that spans departments,
[00:01:32.130]spans data questions,
[00:01:33.870]and she's involved
[00:01:34.710]in this Agricultural Genomes to Phenome Initiative
[00:01:37.050]that she's gonna talk about today.
[00:01:39.270]So with that, it's my pleasure to introduce our speaker.
[00:01:41.940]I have a short bio here.
[00:01:44.040]So Dr. Clarke is a professor
[00:01:45.510]in the Department of Statistics
[00:01:46.800]and Food Science and Technology.
[00:01:49.200]She's additionally the Director
[00:01:50.400]of the Quantitative Life Sciences Initiative,
[00:01:53.310]which serves as the nexus for Data and Life Sciences at UNL.
[00:01:56.850]Sounds very, profound.
[00:01:58.980]Wow.
[00:01:59.813]Even I'm impressed.
[00:02:01.193](Dr. Clarke and Jessica laugh)
[00:02:02.520]She wears many hats.
[00:02:03.480]She's the Chair of the Academic Section
[00:02:05.220]of International Plant Phenotyping Network,
[00:02:08.130]lead the USDA National Agricultural
[00:02:10.260]Producers Data Cooperative
[00:02:12.540]and as I mentioned, you're an Executive Board Member
[00:02:15.120]for the Agricultural Genome-to-Phenome Initiative,
[00:02:17.670]or AG2PI as it's abbreviated.
[00:02:21.300]So, Dr. Clarke's collaborative interest
[00:02:22.950]includes statistical image analysis,
[00:02:24.930]chemical and biological trace evidence
[00:02:26.820]and predictive metagenomics.
[00:02:29.220]You might know her husband, Dr. Bertrand Clarke,
[00:02:31.530]the Chair of the Department of Statistics.
[00:02:33.660]And a personal note, you noted that you're foster parents
[00:02:36.540]to various animals and also parents to human children.
[00:02:39.555]Yes. Four of them. (laughs)
[00:02:41.004](laughing) Yes.
[00:02:41.837]So with that, I'll turn it over to you.
[00:02:43.712]Thank you. Thank you for joining us.
[00:02:44.545]Thank you, Jessica. It was very nice of you.
[00:02:46.278]Okay. Welcome, everyone.
[00:02:49.860]So I was told that today is
[00:02:53.700]as much a discussion as a seminar.
[00:02:55.560]So what I'm going to cover,
[00:02:59.790]or maybe, maybe not get through all of it,
[00:03:01.920]but I wanted to highlight three different national,
[00:03:06.930]international projects that I'm involved in
[00:03:08.760]that I think at least have some relevance
[00:03:11.310]to what Jessica and Dan have been working on,
[00:03:14.940]which is statistical genetics
[00:03:16.410]and also linking the phenotype,
[00:03:19.050]which is an area I've worked in for quite a while.
[00:03:21.060]So I'll talk about the AG2PI project
[00:03:24.990]that USDA is supporting.
[00:03:27.750]I'll talk about our plant phenotyping networks, briefly,
[00:03:31.560]both national and international.
[00:03:33.870]And then I'll talk about a data project
[00:03:38.190]that I'm leading for USDA
[00:03:39.870]that's trying to set up some national cyber infrastructure.
[00:03:43.560]Okay, stop me while I'm speaking.
[00:03:47.040]And welcome to everybody who's online.
[00:03:49.260]If you're online and you have questions,
[00:03:50.970]put them in the chat, and we'll,
[00:03:53.220]or maybe they ask afterwards.
[00:03:54.480]Do they they ask afterwards?
[00:03:56.310]Okay, so in the Q&A ask your questions.
[00:03:59.430]People in the room, if something comes up,
[00:04:01.920]you have a question, just raise your hand
[00:04:03.600]because, you know, we wanna discuss as much as listen, okay.
[00:04:07.710]So this is a project that I'm involved in with colleagues
[00:04:13.890]at different institutions.
[00:04:15.150]And you can see the institutions down here.
[00:04:18.810]We got tasked,
[00:04:21.390]hold on one second.
[00:04:22.530]Maybe I need to stand closer.
[00:04:25.140]Maybe I need to press a different button.
[00:04:27.987](moderator speaking off-mic)
[00:04:31.910]Mute.
[00:04:34.177]Okay, that works. Now it should work,
[00:04:35.160]Now it should work. Oh, fantastic.
[00:04:36.930]Okay.
[00:04:38.767]The next one.
[00:04:39.930]I'll just put all this up at the same time.
[00:04:41.850]Okay, so,
[00:04:47.610]I don't know how many people you are aware of this,
[00:04:48.930]the 2018 Farm Bill directed USDA NIFA
[00:04:53.010]to establish a new competitive grant program
[00:04:57.210]to support research connecting genomes and phenomes
[00:05:00.713]of both the crops and animals, right?
[00:05:03.510]So they authorized up to 40 million annually
[00:05:08.910]from 2019 to 2023.
[00:05:12.540]Everybody went, "Yay."
[00:05:15.030]Except there's a difference between authorization
[00:05:17.400]and appropriation, unfortunately.
[00:05:19.890]And so what they actually appropriated was this,
[00:05:27.207]okay, we're getting somewhere.
[00:05:28.620]It's better than zero.
[00:05:29.760]It's non-zero, definitely non-zero.
[00:05:33.300]So what USDA decided to do was try to build a community
[00:05:41.010]across crops and animals,
[00:05:42.510]just like you guys are starting to do right now
[00:05:44.820]with the joint seminars, right?
[00:05:46.140]So the idea was bring together animal scientists,
[00:05:49.020]and crop scientists, and data scientists and biologists
[00:05:53.100]and get them all to work together
[00:05:55.740]and get them to find out from the community
[00:05:57.870]where USDA should be investing
[00:06:01.890]in order to go back to Congress and say,
[00:06:04.020]hey, you know, that 40 million we were talking about,
[00:06:06.450]we really do need 40 million, not 1 million, right?
[00:06:09.630]40 million a year would be even better.
[00:06:12.900]So a group of us got together and said,
[00:06:18.547]"Okay, we can start working on this."
[00:06:23.220]So this AG2PI project is really designed
[00:06:27.450]to bring scientists together, right?
[00:06:29.220]So it's build the community,
[00:06:31.110]identify the problems that we need to solve,
[00:06:34.710]in some spaces, find out where we stand
[00:06:37.500]in terms of research and translation to practice,
[00:06:41.670]look for gaps, and then report back to USDA.
[00:06:47.460]So kind of feel like a journal,
[00:06:49.530]I don't know, like I'm a journalist
[00:06:50.520]or like I'm a field agent or something.
[00:06:52.050]It's like, "Go out there, find out what the community needs
[00:06:53.640]and then come back and tell us," okay.
[00:06:56.400]So as you can see, it's a group of us.
[00:07:01.916]Pat Schnable's a plant scientist.
[00:07:03.690]David Earl works with a Corn Promotion Board.
[00:07:06.990]Carolyn's a bioinformatician.
[00:07:09.840]Eric's a data scientist.
[00:07:12.690]I'm a statistician.
[00:07:14.760]Then Jack and Chris and Brenda are animal scientist.
[00:07:18.450]So we tried to kind of cover
[00:07:20.880]as much of the space as we could
[00:07:24.210]and then work with all our partners.
[00:07:26.490]So what do we need?
[00:07:28.080]Okay, you need a community.
[00:07:31.539](crackling drowns out speaker) people who usually don't talk
[00:07:33.630]to each other.
[00:07:36.150]You need a lot of data
[00:07:39.090]to solve the genome to phenome link.
[00:07:45.150]I love it when I see this kind of equation,
[00:07:48.750]G by E and P
[00:07:50.250]then people go, "Oh, I'm gonna take my deep hanger
[00:07:52.650]and I'm gonna throw this at this,
[00:07:53.940]and I have 10 samples."
[00:07:54.900]And I go, "You can,
[00:07:59.130]but I don't think you're gonna get anything."
[00:08:02.640]'Cause, you know, as we all know,
[00:08:03.960]solving this problem means you need a lot
[00:08:07.890]of variability, right?
[00:08:09.210]You need variability in your genotypes,
[00:08:10.680]you need variability in your environment.
[00:08:12.420]And then if you wanna do this thing,
[00:08:14.520]okay, the data starts to go up and up and up and up.
[00:08:18.810]So a lot of focus on data
[00:08:21.210]and access to data and understanding what data we have
[00:08:23.850]and what data we need.
[00:08:26.310]And then talk about,
[00:08:28.470]try to tell USDA where to put new investments, right?
[00:08:32.130]So try to get new money into the pipeline.
[00:08:35.550]All right, so this is how the project is structured.
[00:08:39.210]There's the Executive Board, which we don't get the money.
[00:08:44.430]I mean, we get the money, but then we give out the money.
[00:08:46.530]So, unfortunately, we don't get to keep the money. (laughs)
[00:08:49.110]The money kind of flows past us.
[00:08:52.680]We have a Scientific Advisory Board
[00:08:54.750]with people from different countries on it.
[00:08:56.850]We have an External Stakeholder Committee
[00:08:59.760]that gives us guidance.
[00:09:01.950]We have a bunch of partner organizations
[00:09:04.140]that are involved in this space.
[00:09:07.470]Some are animal science, some are data science,
[00:09:10.020]some are plant sciences,
[00:09:12.420]some are cyber infrastructure, right?
[00:09:14.010]So covers a wide space.
[00:09:18.150]And then we have a bunch of member organizations, right?
[00:09:21.810]And the purpose of the stakeholders is communication
[00:09:26.730]to people on the ground,
[00:09:31.560]trying to figure out what they need,
[00:09:33.060]what are the problems they want us to solve.
[00:09:36.150]Give us feedback.
[00:09:37.050]We've done surveys.
[00:09:38.190]I'll cover those a little bit.
[00:09:41.370]They come to meetings,
[00:09:43.020]they tell us what their member concerns are,
[00:09:47.130]because if you're gonna push Congress to fund this area,
[00:09:51.030]you also have to have these people on the table pushing
[00:09:54.780]to get the funding, okay.
[00:09:57.990]So what do we do?
[00:09:59.100]So this project started in 2020.
[00:10:01.850]We had an initial million dollars
[00:10:04.170]and what we proposed was that we would do education,
[00:10:08.940]training, engagement, surveys, meetings,
[00:10:14.730]and we would give out seed grants, right?
[00:10:18.733]The initial, and I'll get to this,
[00:10:20.940]the initial round of seed grants were quite small,
[00:10:24.780]and it was to give to groups
[00:10:27.810]so they could tell us what they wanted us to do, right?
[00:10:31.200]So it wasn't so much to do research.
[00:10:33.660]I know it's always disappointing
[00:10:34.710]because you always get people go,
[00:10:35.767]"Hey, I have this great research project,
[00:10:37.770]can you find my postdoc?"
[00:10:38.700]And it's like, "No, I wish I could,
[00:10:40.770]but really what I need from you is to go talk
[00:10:42.810]to all your friends and tell us where USDA should invest
[00:10:46.800]so that we can get you money so that then you can apply
[00:10:49.260]and then we can pay research projects," right?
[00:10:51.480]So this was kind of like pre-research planning,
[00:10:54.990]is what we were trying to do.
[00:10:57.900]So they gave us initially a million dollars,
[00:11:00.510]then they gave us another million dollars
[00:11:02.280]to do little more and seed grants and all these activities.
[00:11:06.210]And then they gave us another 2 million
[00:11:08.850]to give out in research projects, okay?
[00:11:11.220]So this is now, we're in year three.
[00:11:14.820]So we started out information gathering,
[00:11:17.970]and then we moved to funding small projects,
[00:11:20.850]and then we moved to funding larger projects.
[00:11:23.370]And we just finished this part,
[00:11:25.830]which was giving out close to 2 million in research funds.
[00:11:31.290]Okay.
[00:11:33.000]So what's the goal?
[00:11:36.060]Bring the community together, talk about ongoing research,
[00:11:40.200]talk about resources that are available,
[00:11:43.020]share techniques, share methods, identify gaps,
[00:11:49.980]and then have meetings, including in-person,
[00:11:53.160]which we didn't do during the pandemic
[00:11:56.070]but we're starting to now
[00:11:58.290]to bring people together so they can come up
[00:12:01.770]with shared solutions, right?
[00:12:03.360]So a lot of these seed grants
[00:12:05.190]that we gave were aimed at solving issues
[00:12:09.570]that affect both animal and plant scientists.
[00:12:14.040]And that gets tricky
[00:12:15.030]because most plant scientists
[00:12:16.260]don't know any animal scientists,
[00:12:17.550]and most animal scientists
[00:12:18.570]don't know any plant scientists, right?
[00:12:20.190]So I think for the community it was a little bit of a,
[00:12:24.660]hmm, like I gotta go find a plant person,
[00:12:29.400]or I gotta go find an animal person
[00:12:32.130]to help us on this project, right?
[00:12:34.830]And a lot of things translate,
[00:12:36.480]but not everything translates.
[00:12:37.950]Okay, so one of the things we did was field days
[00:12:40.920]where we highlight ongoing projects in the community,
[00:12:44.610]we talk about different activities.
[00:12:47.070]So these are just examples.
[00:12:48.180]And if you go to the AG2PI website,
[00:12:51.270]all this stuff is posted,
[00:12:52.980]all the slides and the materials, right?
[00:12:56.820]But you can see things that are shared, right?
[00:13:00.570]Genetic genomic prediction and analysis models,
[00:13:03.900]microbiomes, multi-omics,
[00:13:06.840]AI, I mean, a lot of these are tools,
[00:13:11.130]techniques that all of us use, right, regardless.
[00:13:15.210]The way we use them might be a little bit different,
[00:13:17.160]but they apply to both animal plant sciences, okay.
[00:13:22.740]Then we did,
[00:13:25.350]let me skip that,
[00:13:26.183]we did training workshops
[00:13:28.500]where people would sign up online,
[00:13:33.210]and we were all virtual.
[00:13:34.050]Most of them were just a few hours up to a day.
[00:13:38.040]And we tried to give the community basic data skills.
[00:13:45.090]What's Unix?
[00:13:46.620]How do you do SNP analysis?
[00:13:48.960]How do you do GWAS?
[00:13:53.730]What are some ways
[00:13:55.170]to do structural variant detection, right?
[00:13:57.660]So it was aimed at that genome to phenome gap
[00:14:01.470]and basic skills, right?
[00:14:04.080]So if you didn't know anything about these areas.
[00:14:06.150]You could sign on
[00:14:07.110]and you could spend an hour or two doing something,
[00:14:08.700]learning about it.
[00:14:11.460]All of this material is also online.
[00:14:14.130]So it's a good way especially for students
[00:14:17.430]or for faculty who don't know an area
[00:14:19.680]to go, okay, instead of signing up for a course,
[00:14:21.480]it's gonna take me the entire semester
[00:14:24.780]and as a major investment.
[00:14:26.040]And it might not even,
[00:14:26.970]most of it might not even be relevant.
[00:14:28.380]I can just go for a few hours,
[00:14:30.960]I'll learn something about the area,
[00:14:32.610]and then I can better figure out what part
[00:14:36.000]of that discipline fits what my needs are,
[00:14:38.460]and then I can go after something more targeted
[00:14:40.800]that's gonna help me further my research or my studies.
[00:14:44.850]That was kind of the purpose.
[00:14:47.130]Okay, the AG2PI conference coming up.
[00:14:52.170]It's gonna be this summer
[00:14:54.540]in Kansas City co-hosted with USDA.
[00:14:58.260]So USDA people are gonna be there.
[00:15:01.650]And we're gonna talk about what we've done
[00:15:04.110]for the last three years,
[00:15:09.143]what's USDA's next steps.
[00:15:10.800]So right now USDA is saying,
[00:15:14.287]"We're gonna put out an RFP for bigger projects,"
[00:15:17.880]and we're all like, "That's great."
[00:15:20.100]And then that was supposed to be January,
[00:15:21.960]and now it's February, and it's USDA, so.
[00:15:26.700]Maybe sometime by September we'll get something, right?
[00:15:28.800]So it takes time to move these things.
[00:15:32.640]So at this event, we're gonna talk to stakeholders,
[00:15:34.890]we're gonna talk to researchers,
[00:15:36.720]we're gonna have seed grant presentations
[00:15:38.370]and we're gonna talk to USDA
[00:15:40.410]about what are they looking to fund, right?
[00:15:44.190]And what have we learned from this whole process.
[00:15:46.470]So hopefully we get our $40 million.
[00:15:50.070]Okay, we did a survey,
[00:15:53.460]which I thought was sort of interesting.
[00:15:56.640]We asked the community,
[00:15:59.970]what do you think is really important to fund, right?
[00:16:03.990]Where should we be putting our funds, research funds?
[00:16:06.780]And then what's hard to do?
[00:16:08.610]What's easy to do?
[00:16:13.480]And I thought this was actually pretty interesting, so.
[00:16:16.530]So things that are difficult to do
[00:16:20.550]but nobody wants to do is this stuff, right?
[00:16:27.930]Hard.
[00:16:29.550]And if you are primarily a researcher,
[00:16:33.840]maybe not really rewarding,
[00:16:35.790]maybe doesn't get you
[00:16:37.410]the super important first author publication
[00:16:42.331]that you need for your lab, right?
[00:16:45.180]And so it's kind of like, yeah, this is important,
[00:16:47.790]but we don't really wanna put the money there.
[00:16:53.790]We wanna put the money over here, right?
[00:16:58.740]This is the critical stuff that we need to fund
[00:17:02.640]that we think we can actually do as a community, right?
[00:17:07.830]Then there's like easy and not so important.
[00:17:11.670]Clearly most of these people have never tried
[00:17:13.080]to set a standard because you tried to set a standard,
[00:17:15.300]you end up with 15 standards.
[00:17:17.520]So I don't know who thought that was easy, but, okay.
[00:17:21.390]And then you have kind of hard, but we need to do it,
[00:17:26.370]which was phenotyping technology.
[00:17:31.050]I think mostly this was considered hard
[00:17:33.570]and important because we really needed
[00:17:37.890]to capture new phenotypes, novel phenotypes
[00:17:41.040]and it's hard to do that.
[00:17:43.980]A lot of times it involves more complex imaging,
[00:17:46.710]more complex assays,
[00:17:48.750]large data that people really don't know how to deal with.
[00:17:51.600]And so they're like, "Eh, we gotta do it,
[00:17:53.430]but it's kind of tough.
[00:17:54.870]Let's just do this, okay?"
[00:17:58.050]But, you know, this is important for the community to hear,
[00:18:02.910]and this is important for USDA to hear, right?
[00:18:06.060]Is what do we think we can do right now as a community?
[00:18:10.110]If you gave us money, what could we actually achieve?
[00:18:12.930]And is it important? Right?
[00:18:17.190]Okay, so I just wanted to make sure I shared that.
[00:18:21.120]All right, next, oh wait,
[00:18:22.740]let me go back one.
[00:18:24.900]Okay, seed funds.
[00:18:26.280]So we gave out a bunch of seed grants.
[00:18:30.570]You can see that, you know,
[00:18:32.010]the first round were quite small
[00:18:34.230]because it was just, you know, tell us what you need.
[00:18:36.420]It was kind of this,
[00:18:38.130]these are quite small, right?
[00:18:40.380]Just enough for you and your friends to get together
[00:18:43.470]and have cookies, and coffee and maybe beer,
[00:18:46.620]and, you know, talk about if you had a lot of money,
[00:18:49.470]what would you do?
[00:18:51.870]Next round was a little bit larger, right?
[00:18:56.760]Still kind of small, still kind of get things going.
[00:18:59.940]But if you haven't worked with both plant
[00:19:03.810]and animal scientists before,
[00:19:06.030]or you haven't worked with data scientists before,
[00:19:09.090]this is kind of pilot,
[00:19:10.380]you know, this is pilot level funding
[00:19:12.420]where you can start a project
[00:19:14.910]that's kind of in a new area.
[00:19:16.470]You can sit down with people you haven't talked to before.
[00:19:20.760]You don't know if it's gonna work, right?
[00:19:22.170]It's pretty high risk because you're sitting with people
[00:19:24.600]that you may or may not get along with
[00:19:26.010]and you may not have shared interests.
[00:19:29.910]And then we did round three,
[00:19:34.620]which are getting, you know,
[00:19:35.640]they're getting a little bigger, not that big.
[00:19:39.390]And then we just did round four,
[00:19:41.220]which, you know, now, okay, 250K is not chump change.
[00:19:47.010]You know, it's not like winning the lottery,
[00:19:48.990]but it's enough that you could fund some students, right?
[00:19:53.040]You could start to do something more substantial.
[00:19:56.940]You didn't have to be funded in an earlier round
[00:19:59.040]to get the later round, right?
[00:20:00.270]So it wasn't like, oh, I'm sorry,
[00:20:02.040]if you didn't get the seed money,
[00:20:03.240]you know, you're not eligible for the bigger money.
[00:20:05.850]Anybody could apply.
[00:20:10.522]We tried to emphasize cross-kingdom, so animals, plants,
[00:20:14.790]or at least techniques that could be applied in both areas.
[00:20:18.240]And we try to emphasize multi-institutional,
[00:20:25.440]so, you know, work with other colleagues,
[00:20:28.680]you know, work across institutions to solve problems.
[00:20:32.640]I will tell you,
[00:20:35.310]this is being,
[00:20:36.143]yeah, this is being recorded, that's too bad.
[00:20:37.500]Okay, I will tell you that working
[00:20:41.190]across institutions is still difficult.
[00:20:43.290]So for those of you
[00:20:44.123]who have done multi-institutional proposals,
[00:20:47.040]the logistics of doing that can be
[00:20:51.300]sometimes a big wall to try to scale, right?
[00:20:56.070]So, so one of the other things we're telling back
[00:20:58.890]to federal agencies is,
[00:20:59.733]man, you gotta make that process easier.
[00:21:02.340]If you really want institutions to work together,
[00:21:06.600]you know, making the paperwork
[00:21:08.160]even easier would be really nice.
[00:21:10.860]Okay, so these are the types of projects,
[00:21:16.230]and I just, you don't have to read all of it.
[00:21:19.164]I didn't want you to necessarily read all of it,
[00:21:20.370]but these topic areas came from USDA
[00:21:27.960]and from the community.
[00:21:29.550]So this is the kind of stuff
[00:21:32.640]that they're looking to support with additional funding.
[00:21:38.850]So data, cross-fertilization, training,
[00:21:47.790]genome engineering, environmental impacts,
[00:21:52.290]carbon capture.
[00:21:56.220]One of the areas that was, I think,
[00:21:59.100]difficult even for our team to try
[00:22:02.250]to get engagement in was this one, right?
[00:22:06.690]The social, legal, economic aspects.
[00:22:11.850]It's super, super important.
[00:22:16.380]But it was just,
[00:22:17.700]it was tough to engage, right?
[00:22:20.700]It was tough to get those people on board.
[00:22:24.780]So hopefully we'll do better in the future.
[00:22:27.960]Okay, I'm gonna skip this stuff
[00:22:29.580]because I'll put it in the slides,
[00:22:30.600]but it's not really important for me to discuss right now.
[00:22:32.730]But I did wanna mention this one as an example of a project.
[00:22:36.270]So this is a project I'm not involved in,
[00:22:39.930]but I still thought it was pretty cool.
[00:22:42.780]It's trying to encrypt data
[00:22:47.310]so that you can share it and still maintain confidentiality.
[00:22:52.500]And that's a really important problem to solve,
[00:22:56.340]because, you know, as researchers we wanna publish,
[00:22:58.590]and a lot of the publication venues
[00:23:00.000]say you must share your data,
[00:23:01.620]but a lot of the places where we get our data,
[00:23:03.183]they don't wanna sharing our data,
[00:23:05.130]they wanna solving the problem, right?
[00:23:07.020]But they don't necessarily want the data public.
[00:23:11.310]And so I thought this was a cool idea
[00:23:14.220]that you could actually figure out a way
[00:23:17.430]to still satisfy the requirements of journals
[00:23:19.800]for sharing your data
[00:23:21.240]in the sense that if you analyze the data that's provided,
[00:23:24.540]you'll get the same answers as what's shown in the paper.
[00:23:28.980]But the data's been encrypted,
[00:23:30.660]so it's not actually the data that was originally analyzed,
[00:23:34.200]but when you do a homomorphic translation,
[00:23:38.160]the relationships in the data stay the same,
[00:23:40.170]the values change, the relationships stay.
[00:23:43.440]So it's kind of a cool idea.
[00:23:46.920]And it came out because of these things, right?
[00:23:50.130]That there's a big push to share data FAIR standards.
[00:23:55.542]I don't know if anybody's remember familiar
[00:23:56.700]with FAIR, findable, accessible,
[00:23:58.890]inter-operatable, reproducible research.
[00:24:03.360]Huge pressure coming down from federal agencies
[00:24:07.170]for everyone to share the products of their research.
[00:24:11.580]And that's all well and good,
[00:24:13.500]unless your data is under an NDA
[00:24:16.860]or your data comes from a private provider who says,
[00:24:19.860]look, I want you to work with my data,
[00:24:21.360]but it's not public.
[00:24:23.940]I don't want it share it, right?
[00:24:27.060]So it's trying to satisfy this requirement
[00:24:31.440]and also satisfy the requirement of privacy, right?
[00:24:35.070]So that's particularly industry partners feel comfortable
[00:24:37.470]working with you and knowing that you're not gonna have
[00:24:41.790]to make everything public, okay.
[00:24:44.520]So this is ongoing.
[00:24:46.680]There are funds still available
[00:24:49.800]for setting up working groups
[00:24:53.580]and doing like one-off projects.
[00:24:56.790]So if you wanna host a workshop,
[00:24:59.310]or a conference or a short course, there's funds to do that.
[00:25:05.250]You can apply.
[00:25:07.050]If you wanna put together a group of colleagues in your area
[00:25:10.980]and, you know, start a community focused on something
[00:25:14.400]that's relevant to this space,
[00:25:16.080]then you can request funding for that.
[00:25:19.710]Just so you know. All right.
[00:25:22.650]Questions, comments?
[00:25:27.360]Yes. I have a question
[00:25:28.193]regarding the 40 million.
[00:25:29.940]Yep.
[00:25:30.773]That could happen.
[00:25:31.770]Yeah. (laughs)
[00:25:32.730]Would that still be brokered
[00:25:34.530]through your executive board,
[00:25:36.016]or will that be- Oh God, no.
[00:25:36.849]Oh my God, (laughing) I hope not.
[00:25:39.990]I mean, that's a very good question.
[00:25:46.110]And that's, to be honest,
[00:25:48.330]that's one that we don't really wanna do, right?
[00:25:51.870]Because how this turned out was
[00:25:56.325]because they continued to give us additional funds
[00:25:59.160]and then wanted to give them out as grants,
[00:26:01.020]and I was like, "that's great,"
[00:26:03.660]we start feeling like program managers and it's like,
[00:26:06.570]yeah, "I wanna help the community,
[00:26:08.280]but I'm not a program manager."
[00:26:12.180]And so right now what USDA is telling us
[00:26:16.140]is that they're gonna put out,
[00:26:18.090]they're gonna manage it, right?
[00:26:19.260]At this point we're like, this is our last round.
[00:26:22.530]We've spent our $4 million
[00:26:24.630]and we feel like the community's in a really good spot
[00:26:27.510]and it's kind of time for a significant investment, right?
[00:26:30.810]And that's USDA's job,
[00:26:33.510]is to put that out, and put out the RFA
[00:26:36.150]and manage all the submissions.
[00:26:37.380]So (laughing) hopefully.
[00:26:40.890]If not, I think there's a lot of people here
[00:26:42.690]that like to take you out for beer. (laughs)
[00:26:44.178]Oh, yeah, I know, I know,
[00:26:45.600]and that's why I was like,
[00:26:46.433]"How should I answer this question?"
[00:26:47.400]No, I should answer it honestly.
[00:26:49.560]But, you know, the point is we don't really know right now.
[00:26:53.250]It's just we're kind of hopeful that,
[00:26:55.673]you know, they take it over.
[00:26:58.590]And part of that is because US on the executive team,
[00:27:01.590]you know, we're all researchers,
[00:27:02.820]so because the money's coming to us,
[00:27:05.760]we're not eligible, right?
[00:27:07.320]We have to step out and say, we're not doing it.
[00:27:11.559]There's a conflict of interest.
[00:27:13.080]We are not gonna be eligible for any of this funding.
[00:27:17.070]But, you know, we got into this
[00:27:18.330]because we wanna do the research like everyone else.
[00:27:21.450]So I would much rather have somebody else take applications
[00:27:25.140]and then be able to apply.
[00:27:27.060]Yeah.
[00:27:28.624]Go back.
[00:27:30.060]Yeah.
[00:27:32.790]This one?
[00:27:33.947]Yeah.
[00:27:34.780]Uh-huh?
[00:27:35.613]Isn't that the public fund?
[00:27:40.430]Yeah, so the funds go to UC Davis.
[00:27:46.350]Yeah, but (speaking off-mic).
[00:27:48.690]Yeah, they're advisors,
[00:27:50.460]but they don't get any money.
[00:27:53.040]There's no money going to them, right?
[00:27:55.080]So yes, I agree that if they were receiving any funds,
[00:27:59.940]then I would say yes, it's a conflict.
[00:28:02.430]And I'll tell you that they were not involved in the review
[00:28:05.760]of these because they have to step off
[00:28:08.370]even if they're unfunded collaborators.
[00:28:12.210]But they're not getting any funding out of it, right.
[00:28:17.700]While you're here,
[00:28:18.888]I don't think this is any secret.
[00:28:20.190]So Tianjing, she goes by TJ,
[00:28:22.590]she will be joining our faculty in August
[00:28:25.950]as quantitative geneticist.
[00:28:28.620]Yeah, I must have swami powers
[00:28:32.190]'cause like, I didn't know that,
[00:28:33.300]and I'm like, "Now I totally nailed it
[00:28:35.070]by the project I picked."
[00:28:35.970]That's awesome.
[00:28:37.702](laughs) So happy.
[00:28:38.535]Okay. Good questions.
[00:28:40.290]Good questions.
[00:28:42.450]Anything else before I move on?
[00:28:45.660]Okay. Okay.
[00:28:48.540]Okay, next, next, next, next.
[00:28:50.040]Where are we here? Ah, okay.
[00:28:51.510]So for the plant people in the room,
[00:28:55.170]and the animal people, you should still pay attention.
[00:28:57.180]But for the plant people,
[00:28:58.410]I'm gonna do just a few slides
[00:29:01.560]on the North American Plant Phenotyping Network
[00:29:06.000]and the international one, right?
[00:29:07.230]Just so you're aware, these organizations exist.
[00:29:10.380]So the North American Plant Phenotyping Network
[00:29:15.690]started in 2016.
[00:29:18.540]It wasn't even a network then.
[00:29:19.740]It was just a bunch of people getting together going,
[00:29:21.727]"Wow, we should have a network.
[00:29:23.610]We should share all this information."
[00:29:29.512]It's a group of scientists,
[00:29:30.510]researchers who are interested in plant phenotyping.
[00:29:34.470]So as you're probably aware,
[00:29:36.690]there's a big greenhouse sitting out at Nick
[00:29:38.940]that has all those fancy conveyor belts in it.
[00:29:41.820]And probably if you pay Vincent enough money,
[00:29:45.030]he'll let you sit in the hyperspectral chamber
[00:29:48.001]or you can sit in the thermal chamber
[00:29:48.840]and get your picture taken.
[00:29:51.990]So when a lot of those technologies hit the plant sciences,
[00:29:56.910]they didn't know, they're like, "What the?
[00:29:59.010]I don't know how to deal with this stuff.
[00:30:00.600]I mean, I don't know how to do hyperspectral.
[00:30:03.300]I don't know how to do fluorescent imaging.
[00:30:08.460]How do we even use these technologies in research?"
[00:30:11.790]And so a bunch of us got together and said,
[00:30:14.827]"Well, all right, we should build a network, right?
[00:30:18.390]Share information."
[00:30:19.223]So that's where this organization came from.
[00:30:23.340]It's very multidisciplinary, which is really cool.
[00:30:28.170]It's nonprofit.
[00:30:30.870]We have about 300 members of all different ranks,
[00:30:36.120]positions, backgrounds, industry,
[00:30:39.870]commodity boards, academics.
[00:30:43.650]I mean, it's like, it's just everybody.
[00:30:46.830]And it's super fun.
[00:30:50.380]And they put a ton of information on their webpage.
[00:30:54.060]They have six affinity groups,
[00:30:56.700]which is where a lot of the action happens, right?
[00:30:59.730]So if you're new to the area and you're like,
[00:31:03.570]I need to learn this stuff,
[00:31:04.890]or you're a student
[00:31:06.180]and your advisor says you need to learn this stuff,
[00:31:09.540]this is a good place to go.
[00:31:11.970]It's very open.
[00:31:13.530]They have a Listserv, you can send them messages.
[00:31:18.360]Super inexpensive to join.
[00:31:21.030]We have a conference every year.
[00:31:22.650]We just had one at the Danforth in St. Louis.
[00:31:27.300]And, you know, students give talks,
[00:31:32.280]early career people give talks, we give awards,
[00:31:35.400]we have poster sessions
[00:31:38.250]and we do a lot of interacting with industry
[00:31:40.260]because a lot of the really expensive phenotyping systems
[00:31:43.620]are not at academic institutions,
[00:31:46.200]they're an industry, right?
[00:31:49.890]So encourage people to contact and get involved
[00:31:52.830]'cause it's super cool.
[00:31:55.410]And then one of the things that I wanted to,
[00:31:58.290]sorry, that I wanted to highlight was
[00:32:00.060]that the Europeans are actually doing much more
[00:32:06.000]in plant phenotyping than we are,
[00:32:08.520]and started in it,
[00:32:09.720]actually the Australians were probably the first ones
[00:32:11.610]to do it on a large scale.
[00:32:14.910]So we as the University of Nebraska are a member
[00:32:18.900]of the International Plant Genotyping Network,
[00:32:21.180]which functions differently.
[00:32:23.070]It's institutional members,
[00:32:25.290]it's not individual members, right?
[00:32:26.970]So, you know, you can't just contact them and sign up.
[00:32:30.510]Your institution has to be a member, right?
[00:32:32.760]So we are a member,
[00:32:35.910]which means we have connections with the Europeans,
[00:32:38.700]and Australians, and the Brazilians
[00:32:40.380]and the people in Sub-Saharan Africa
[00:32:43.980]who are trying to do phenotyping,
[00:32:45.600]affordable phenotyping in the field.
[00:32:48.420]So if you want to, you can go to their webpage.
[00:32:52.260]These are all their members.
[00:32:55.800]And it's a great way to get connected
[00:32:58.380]with the international community, so.
[00:33:00.420]Their last meeting was last fall
[00:33:02.970]in Bogin in the Netherlands.
[00:33:07.469]And we're trying to figure out
[00:33:08.460]where our next meeting's gonna be.
[00:33:09.810]It's supposed to be in the America somewhere,
[00:33:11.490]but we don't know where yet.
[00:33:12.660]Okay, so just in case you're interested in that space,
[00:33:16.200]trying to connect with researchers in those areas.
[00:33:20.670]Okay, let me skip this one.
[00:33:23.100]Okay, and they have working groups too.
[00:33:25.950]Here are their working groups.
[00:33:30.510]The Forest one is really cool.
[00:33:32.970]It's one that the North Americans, we don't have.
[00:33:37.800]They have a seed in germplasm group.
[00:33:40.590]That's a little different.
[00:33:43.650]So it's just a broader community, right?
[00:33:46.980]Because some of the activities are just more active
[00:33:49.800]outside of North America.
[00:33:51.750]So if you wanna talk to people who actually do a lot
[00:33:53.550]of this stuff, you talk to the Europeans,
[00:33:56.370]which is interesting 'cause you can end up in meetings
[00:34:00.030]where you have to get up really early in the morning,
[00:34:01.560]I'm sorry to say that,
[00:34:04.173]'cause if you got meetings with the Europeans
[00:34:05.610]and the Chinese, that means you're up super early,
[00:34:09.810]the Chinese are up super late
[00:34:11.400]and the Europeans are laughing at both of you
[00:34:12.377]'cause it's two in the afternoon there.
[00:34:14.682](laughs) So bring your coffee.
[00:34:17.190]Okay, I wanted to highlight this meeting,
[00:34:20.820]back one more, (humming)
[00:34:23.400]ah, this meeting.
[00:34:26.370]So there is a USDA group
[00:34:31.860]called the NCCC170.
[00:34:35.220]It's a multi-state hatch project on agricultural statistics.
[00:34:40.470]And since I'm statistics,
[00:34:43.470]and I know that all the stuff you do involves statistics
[00:34:46.800]at some point, then this is a good meeting to go to, right?
[00:34:53.310]It's inexpensive.
[00:34:56.130]It's a bunch of people who do agricultural statistics.
[00:35:00.210]So unlike some other statisticians,
[00:35:02.700]they will actually understand what you're talking about.
[00:35:05.236]Right, if you're saying,
[00:35:06.069]talking about plants and animals and genome selection,
[00:35:10.080]they'll be like, "Uh-huh," right?
[00:35:11.910]They understand, they work in this area,
[00:35:16.650]they do workshops and they get speakers
[00:35:19.170]and it's small and you get to interact with them.
[00:35:21.330]So super cool.
[00:35:22.980]And their abstract submission thing is still open.
[00:35:27.690]So you can do a poster.
[00:35:29.340]And it's at Purdue, so it's not that far away.
[00:35:34.633]I just wanted to highlight that
[00:35:35.466]because the meeting is not just statisticians, right?
[00:35:37.770]It's statisticians who are involved in agriculture
[00:35:40.620]and it's also other people working in agriculture
[00:35:43.740]who have to involve statisticians in their projects, right?
[00:35:46.560]So it's more multidisciplinary than just going
[00:35:48.750]to a STAT conference, which I would enjoy,
[00:35:51.870]but, you know, it's not necessarily for everyone, okay.
[00:35:56.160]Okay, so let me talk about this project.
[00:35:58.410]How much time do I have? Oh, I'm still okay.
[00:36:02.550]Okay, so USDA in (humming) 2020,
[00:36:10.530]2020, late 2020,
[00:36:17.370]the federal government at that time suddenly realized
[00:36:20.730]that artificial intelligence is a thing,
[00:36:25.950]yeah, I know, I would laugh too,
[00:36:28.530]but it's getting to be a big thing
[00:36:30.750]and they don't know what it is
[00:36:33.240]and that they have a lot of data.
[00:36:38.850]They don't know where it is or who has it.
[00:36:42.450]But it'd be really nice if they could put all the data
[00:36:45.960]in places where they might be able
[00:36:48.690]to leverage whatever AI does, okay?
[00:36:52.770]And I think a lot of this emphasis came partly
[00:36:57.600]from high profile projects like IBM Watson,
[00:37:01.500]if you're familiar with that,
[00:37:03.420]AlphaGo, which is Google's AI arm.
[00:37:08.640]It was kind of this sudden awareness that, oh my gosh,
[00:37:10.470]like, you can do kind of cool
[00:37:12.810]and also sort of scary things with ai.
[00:37:16.380]And so they went to each of their research agencies
[00:37:20.670]and said, "You guys have a lot of data,
[00:37:21.960]like, we've been putting a lot of tax money
[00:37:23.730]into what you guys are doing.
[00:37:25.440]Where's the data?
[00:37:26.760]So we have these AI people that wanna do cool stuff on it.
[00:37:30.386]And then NIH went, "Yeah, okay, we have NCBI
[00:37:33.777]and that's where we put a lot of the data
[00:37:35.280]and that's where we tell our researchers to put data."
[00:37:37.808]And they went, "Um-huh, that's good."
[00:37:39.536]And then NSF they went, NSF said,
[00:37:41.245]"Well, we have a data management plan
[00:37:42.824]and we know where a lot of the data is.
[00:37:46.312]And they went, "That's good."
[00:37:47.426]And then they went to USDA and USDA went,
[00:37:52.027]"Ah, I dunno, it's somewhere.
[00:37:55.410]It's lots of places," right?
[00:37:59.040]And I think honestly, USDA probably has at least
[00:38:01.890]as much data as those other agencies,
[00:38:03.540]but it's just not as organized.
[00:38:07.470]And partly to be fair, I think that's in large part
[00:38:10.800]because the data that USDA researchers work with
[00:38:13.710]is very diverse,
[00:38:17.130]at least as diverse as the other agencies,
[00:38:19.410]and can be very large, right?
[00:38:22.710]And USDA, you know, they don't have a data farm
[00:38:26.040]out in the middle of somewhere where they can just go,
[00:38:28.890]oh, you have five terabytes of data, no problem.
[00:38:31.410]Just, you know, just send it over, right?
[00:38:34.410]They don't have that capacity.
[00:38:36.090]So what USDA decided to do was, "Okay,
[00:38:43.170]clearly we need to change this scenario.
[00:38:46.860]We don't have a money tree,
[00:38:49.890]so why don't we,
[00:38:53.520]we'll task a group to come up with a framework
[00:38:58.200]on the national level that would enable us
[00:39:01.890]to leverage all the land grant institutions,
[00:39:05.400]and all our extension agents
[00:39:07.500]and all the researchers, and they'll tell us what to do.
[00:39:12.870]Okay.
[00:39:13.770]And then all the commodity groups
[00:39:17.520]and producer communities said like,
[00:39:22.650]producers need this, right?
[00:39:24.560]At the time, there was a big debate going on
[00:39:27.330]about right to repair.
[00:39:28.890]I don't know if people were familiar with that,
[00:39:30.990]but basically producers being worried
[00:39:34.470]about not having control of their own data,
[00:39:37.200]that a lot of the data
[00:39:38.130]from their operations were going to companies,
[00:39:41.100]and it was like, I don't know where it goes.
[00:39:44.100]I don't know if it has any value.
[00:39:46.590]You know, it just kind of disappears.
[00:39:49.410]So there was a lot of concern among producer communities,
[00:39:54.240]what's going on with my data, right?
[00:39:57.677]So USDA said, "Okay, okay, okay,
[00:39:59.880]we'll target it to serve producers
[00:40:03.810]and then maybe we can also leverage it to help researchers."
[00:40:06.900]And they were okay.
[00:40:09.210]So this is the group that, you know,
[00:40:12.570]drew the short straw. (laughs)
[00:40:15.142]And so we said,
[00:40:15.975]"Okay, I guess we could try doing something with this"
[00:40:19.920]So we got together in early 2021 and said,
[00:40:24.607]"Okay, this is what we've been tasked to do.
[00:40:27.780]We don't have to build it.
[00:40:30.630]We just have to come up with a framework
[00:40:34.260]that we would follow if we had to build it," okay.
[00:40:38.610]And so what what this cooperative does is try to look
[00:40:44.880]at the data-driven ag space.
[00:40:50.130]It covers genetics, genomics,
[00:40:53.790]it covers phenotyping,
[00:40:55.110]it covers weather, it covers remote sensing,
[00:41:00.360]it covers equipment.
[00:41:05.100]It's just a ton of data.
[00:41:06.673]There's a ton of data.
[00:41:07.980]And what we needed to ask them was,
[00:41:10.830]well, what do you guys need?
[00:41:14.040]Right?
[00:41:15.030]Everybody talks about data-driven ag.
[00:41:17.130]The problem is that that's not necessarily easy to do.
[00:41:22.860]It can be expensive, it can be hard to manage.
[00:41:28.260]There's a lot of barriers to get over.
[00:41:31.830]I've had conversations with producer groups where I say,
[00:41:37.327]"We're doing this project.
[00:41:39.360]I really value your input.
[00:41:41.880]Could you tell us what your community needs
[00:41:44.400]around the data space?
[00:41:45.750]What would enable you to be more productive?"
[00:41:48.900]And the first sentence outta their their mouth is,
[00:41:51.667]"Why is USDA all of a sudden interested in my data?"
[00:41:54.997]"Like, because we wanna help you?"
[00:42:03.060]It's a difficult space.
[00:42:04.800]It's a trust issue for a lot of people, right?
[00:42:09.810]They hear data and they think cybersecurity, privacy.
[00:42:15.630]It makes them nervous, and I respect that.
[00:42:21.360]So what we decided was the idea was
[00:42:24.570]to give them control over their data, right?
[00:42:26.610]So that they don't have
[00:42:27.570]to necessarily use commercial products,
[00:42:29.520]they don't necessarily have to give up all their data.
[00:42:32.430]They can use it themselves if they want to.
[00:42:35.940]Okay, so how do you kind of enable that?
[00:42:39.630]Well, we thought first thing to do is identify the needs
[00:42:42.960]and the challenges.
[00:42:44.730]It's physical infrastructure, it's internet,
[00:42:47.640]it's human (laughing) resources, training, education.
[00:42:53.700]It's getting people to engage,
[00:42:55.650]which can be tricky, especially during a pandemic
[00:42:57.690]when nobody wants to engage, I mean.
[00:43:02.460]Fund some pilot projects.
[00:43:05.544]So identify some gaps that we have and try to fund it,
[00:43:08.910]and then tell everybody what we're doing.
[00:43:11.910]So we came up with a list,
[00:43:15.000]and this isn't exhaustive
[00:43:17.310]and you don't have to read all of it,
[00:43:19.500]but we kind of broke up the cyber infrastructure space
[00:43:24.270]into things that are shared across communities, right?
[00:43:29.040]That everybody's gonna have to face.
[00:43:31.500]Data governance, right?
[00:43:34.860]Cyber security, training.
[00:43:39.150]Everybody wants their systems to be sustainable
[00:43:41.670]and reliable, right?
[00:43:43.320]When you turn on your Netflix, you want it to work.
[00:43:45.480]You don't wanna have to sit there buffering, right?
[00:43:48.600]Those are things that everybody is gonna be facing.
[00:43:51.210]And then there's the things
[00:43:52.050]that are kind of domain specific, right?
[00:43:55.530]The specific questions you ask your data are gonna vary
[00:43:59.100]depending on your operation, right?
[00:44:01.470]You might wanna know, okay,
[00:44:03.180]how much fertilizer do I need to use?
[00:44:05.310]And you might wanna ask, how much fee do I need to buy?
[00:44:07.740]Right, different types of questions, different queries.
[00:44:12.120]And you might have different data
[00:44:14.370]that you wanna integrate, right?
[00:44:15.690]So your data sources might be slightly different.
[00:44:20.190]And so we kind of broke it into these two different areas.
[00:44:24.600]And then we said, "Okay, we gotta share information,
[00:44:27.660]so we started a webinar series, which goes every month.
[00:44:31.560]And these are up on our webpage,
[00:44:33.390]the recordings for all the webinars.
[00:44:35.310]And we get people from the plant community,
[00:44:38.010]the animal community,
[00:44:38.940]we get people from National Ag Library,
[00:44:42.150]we get people from universities.
[00:44:45.840]So it's a whole different group of individuals
[00:44:48.750]that come and give webinars and talk to us.
[00:44:52.800]Our next one, actually,
[00:44:53.633]this one's been rescheduled for April,
[00:44:55.770]but we also got the Canadians coming in
[00:44:59.100]to talk about machine intelligence in agriculture.
[00:45:04.320]We gave some awards out, right?
[00:45:05.850]So we had a initial grant mechanism where we gave money
[00:45:09.630]to people and we said,
[00:45:12.307]"You know your area better than we do
[00:45:14.280]so go out, ask your friends,
[00:45:16.080]ask your colleagues what we need to be doing," right?
[00:45:19.140]So we have one here at UNL with Ben Jewel
[00:45:23.790]who's working with specialty crops.
[00:45:25.830]But we have one on data and data sharing.
[00:45:29.340]We have one on working across crop,
[00:45:32.130]livestock and aquaculture.
[00:45:35.010]We have one aim at crops and,
[00:45:39.600]you know, sharing information,
[00:45:40.830]setting up an independent
[00:45:42.480]kind of community owned data storage and sharing system,
[00:45:46.920]which smaller than what John Deere has,
[00:45:50.310]but it's a start, right?
[00:45:51.510]Because it's controlled by their co-op, right?
[00:45:54.660]So they decide what happens with the data,
[00:45:56.700]whether they wanna sell it or not sell it,
[00:45:58.530]what they wanna do with it, okay?
[00:46:01.410]So we do that stuff.
[00:46:04.200]We have a meeting here,
[00:46:06.840]well, not exactly right here, but there,
[00:46:11.340]I think, over there in May.
[00:46:15.090]So you guys will still be here, right?
[00:46:16.727]'Cause the semester runs later this year.
[00:46:19.710]Okay, so come to the meeting.
[00:46:23.910]We're gonna talk about all these issues:
[00:46:27.090]data sharing, data privacy.
[00:46:30.090]We'll talk about
[00:46:34.770]how to work within industry systems.
[00:46:36.480]So people will be here from Deer and Climate Corp.
[00:46:43.140]And it's really to talk about data in agriculture.
[00:46:46.050]So that's a big,
[00:46:47.760]I mean, when you say data in agriculture,
[00:46:49.608]that's a huge topic, right?
[00:46:51.540]But the idea is to have actual discussions
[00:46:54.960]about what we need to build, right?
[00:46:56.400]So we can turn around to USDA and go, please build this.
[00:47:00.960]And they'll say, I don't have any money to build that.
[00:47:03.057]And I'll be like, well, but this is what you need
[00:47:05.940]to be working on building, right?
[00:47:07.320]And it's not a six month project,
[00:47:09.870]it's not a year, it's not probably three to five years,
[00:47:12.510]it's a 10 to 20 year proposition, right?
[00:47:17.610]Okay, I think,
[00:47:20.347]is that that all I have?
[00:47:21.450]I think, yeah, yeah.
[00:47:22.800]So, okay, so you can go sign up.
[00:47:24.870]We actually have a lister, see right there.
[00:47:28.080]You can go contact and sign up
[00:47:30.480]and we'll tell you about webinars,
[00:47:31.980]and events, and meetings and,
[00:47:35.730]you know, anything ag data related, okay.
[00:47:42.690]Are there any, okay, questions from the room first?
[00:47:46.683]Questions, questions, question.
[00:47:48.600]Yes.
[00:47:49.433]We'll pass the mic around here-
[00:47:51.360]That way people online
[00:47:52.230]can hear us. Yes.
[00:47:54.832]And here it is.
[00:47:58.170]Jennifer, thank you for the overview.
[00:48:01.320]Very informative.
[00:48:03.780]My question is, maybe at the local level,
[00:48:07.080]what I see and what the challenges are
[00:48:10.110]for faculty is this hub of talented people
[00:48:14.820]that can process data.
[00:48:17.010]And, you know, I think there's a balance
[00:48:20.010]between collecting the data but also establishing a hub
[00:48:23.310]of data processing, and archiving and pipeline.
[00:48:27.390]Can you share with us what your thoughts around that
[00:48:30.780]and what UNL and INR in particular can think about?
[00:48:35.820]Yeah, that's a fantastic question.
[00:48:40.013]And you're absolutely right,
[00:48:41.370]is we have this gap
[00:48:46.590]between the data that we collect
[00:48:50.760]and getting information out of it.
[00:48:55.410]And that process can be quite complicated
[00:48:59.910]and quite challenging.
[00:49:02.424]And I think what we've done to date is we say,
[00:49:07.620]well, you can use core facilities,
[00:49:12.957]and we do have some core facilities.
[00:49:14.550]We have the bioinformatics core,
[00:49:15.930]we have phenotyping facilities.
[00:49:18.060]So we do have some people
[00:49:20.160]in certain places where we've said,
[00:49:22.320]okay, there's a whole group of researchers that need this,
[00:49:26.070]and so let's have some kind of service center
[00:49:28.950]that might help them work through that.
[00:49:32.430]But as we know from working with service centers and cores,
[00:49:37.740]they're designed to be a one size fits all.
[00:49:40.440]I mean, that's kind of how they work.
[00:49:41.580]And if you want them to do something more specialized,
[00:49:43.950]then it becomes I gotta hire somebody,
[00:49:45.690]I mean, I have to pay part of their time, right?
[00:49:48.737]And so you can get the basics
[00:49:51.000]and you can even get some training,
[00:49:53.610]but a lot of times addressing specific research questions
[00:49:57.570]becomes, "Well, my student is gonna have to figure that out,
[00:50:01.440]or my postdoc is gonna have to figure that out."
[00:50:03.210]And then that postdoc or student graduates,
[00:50:06.510]and then you keep repeating the process, right?
[00:50:08.460]You're like, "Oh,
[00:50:10.050]you're not gonna graduate soon, are you?" (laughs)
[00:50:12.960]And then, "Oh, okay, they're supposed to graduate,"
[00:50:14.970]and then it's like, "Oh no,
[00:50:16.590]my new person doesn't know any of this yet."
[00:50:19.085]Oh, okay, and then we start over.
[00:50:20.730]So in some sense that fulfills our mission;
[00:50:23.670]we're supposed to train a lot of people,
[00:50:25.500]but on the other hand,
[00:50:26.333]it makes it challenging to have a sustainable program
[00:50:28.380]because you're always having the gap fill, right?
[00:50:30.960]And you're like,
[00:50:31.793]"Oh, that person finally gets to know what they're doing,"
[00:50:33.300]and then they're gone, they're gone, you start over again.
[00:50:37.920]I think you're right that we need to hire some people
[00:50:43.110]who want to do that data processing part.
[00:50:47.340]Not everybody wants to do it, right?
[00:50:49.260]Some of them are like, "Ugh, I have to process the data?
[00:50:51.660]Ugh, I don't wanna do that," right?
[00:50:55.620]So you have to find the right,
[00:50:58.662]I can't believe I'm using this word,
[00:50:59.495]you have to use the right phenotype of person who says,
[00:51:02.820]I know it's not weird.
[00:51:04.020]You have to pick people who have some good skills.
[00:51:09.990]They probably want some more experience.
[00:51:12.750]You're still gonna have the heavy turnover.
[00:51:14.580]Because every time we train somebody in the data space,
[00:51:18.570]Bayer, Corteva, Google,
[00:51:21.330]Apple, IBM, I mean, you know,
[00:51:24.150]we'll, (swooping) you know, swoop,
[00:51:26.460]they're gonna swoop and take them
[00:51:27.450]and we can't compete with those kind of salaries.
[00:51:30.240]I mean, no way.
[00:51:33.270]So you want people who've somehow feel like being
[00:51:37.650]in that data processing role
[00:51:39.660]is something that's gonna further their career,
[00:51:41.940]it's going to give them skills
[00:51:43.620]that then they can go out and go,
[00:51:45.420]not only am I trained to do this,
[00:51:47.580]but I've actually done it, right?
[00:51:49.230]I know how to work in teams.
[00:51:51.030]I know how to communicate my data skills.
[00:51:54.270]And then you'll have to have,
[00:51:56.940]I think you really will have to have a group
[00:51:59.160]of people doing that because they're gonna leave
[00:52:02.580]and then you're gonna hopefully have some more senior people
[00:52:05.670]in the group to train the more junior people, right?
[00:52:08.250]So it's kind of a flow through
[00:52:10.830]where you hope that they stay long enough
[00:52:12.660]that they can contribute.
[00:52:13.800]And I think grad students can,
[00:52:18.210]I mean, students can play a role in that space,
[00:52:20.910]but they need a mentor, right?
[00:52:22.620]So you don't wanna just be throwing students in there
[00:52:25.110]and going, hey, take this data and go there with it,
[00:52:28.230]because it's painful,
[00:52:30.780]it's very time consuming, it can be frustrating.
[00:52:34.500]So they need some guidance.
[00:52:36.390]So I kind of imagine you could have a small group
[00:52:38.820]of people who that's their job
[00:52:40.830]is to work with researchers and process data
[00:52:44.010]and set up some pipelines.
[00:52:46.500]You know, there are a lot of pipelines out there, right?
[00:52:49.080]But they're not one size fits all pipelines, right?
[00:52:51.840]So you might have to modify them a little bit
[00:52:53.760]depending on what kind of data we're collecting here.
[00:52:57.660]But that could go a long way because that process,
[00:53:01.560]if you're not familiar, can be a time sink.
[00:53:04.560]And you need to publish, right?
[00:53:07.050]You need to get results,
[00:53:08.580]you need your students graduating,
[00:53:10.080]you need to get your grants.
[00:53:11.040]And so I think a group
[00:53:12.840]that can enable that would be super helpful.
[00:53:16.170]And I've seen that at other institutions.
[00:53:17.820]I've seen them hire people
[00:53:19.110]who are research scientist kind of positions.
[00:53:23.700]They're not faculty, they're not student,
[00:53:27.090]they're literally like data scientist
[00:53:30.630]and that's their job, right?
[00:53:33.090]And if we can leverage that part,
[00:53:35.430]because I know that the students we graduate initially,
[00:53:39.390]yeah, they can get the job at Facebook or whatever,
[00:53:43.230]and they'll say, yeah, we'll pay you 150K
[00:53:45.930]and you can live in the Bay Area,
[00:53:48.000]and you can live in a pup tent and eat ramen
[00:53:51.330]because that's really what you can afford with 150K.
[00:53:54.450]If they want something, a more advanced position,
[00:53:57.570]that's what they really want.
[00:53:58.410]They want something more advanced in those companies.
[00:54:00.720]But in order to be eligible for that,
[00:54:02.970]you gotta have some more experience, right?
[00:54:04.650]You gotta have a postdoc
[00:54:05.550]or you gotta have some work experience.
[00:54:09.180]I think a group here could give them
[00:54:11.070]that work experience where then they could say,
[00:54:14.460]yeah, I know 150 sounded great when I was a grad student,
[00:54:17.220]but now I know 150 doesn't get me very far.
[00:54:19.920]I want the 300, right?
[00:54:23.795]And they'll be in a good position to get it, yeah.
[00:54:30.420]Yeah.
[00:54:31.860]So in the conversation you are having
[00:54:36.990]at least with USDA,
[00:54:38.550]are there talk about capacity building
[00:54:40.860]and grants that specifically (crackling drowns out speaker)
[00:54:44.160]and establish this, you know,
[00:54:46.470]these talents and then, you know,
[00:54:48.780]the institution can take over.
[00:54:50.850]Is that conversation happening?
[00:54:52.170]Yeah. Yeah.
[00:54:53.190]I think that we've started having that conversation
[00:54:57.030]with USDA that the more we talk to them
[00:55:00.840]about the linking genome to phenome,
[00:55:05.190]they gonna go, well, that sounds great.
[00:55:06.870]And then we go, well, in order to do that,
[00:55:09.543]that means we have to have conversations about data.
[00:55:12.570]And (laughs) what's interesting is that most people,
[00:55:17.700]they're not gonna get excited about data,
[00:55:19.650]they're not gonna get excited about pipelines.
[00:55:21.600]I mean, to me, I get excited about this stuff.
[00:55:23.550]I'm like, "Oh, you got a new pipeline?
[00:55:25.200]Let's go see.
[00:55:26.400]Gotta check it out."
[00:55:27.270]Get excited, right?
[00:55:28.110]It's a flow chart.
[00:55:28.943]I'm like, "Whoa."
[00:55:30.150]But USDA is not gonna get excited about that.
[00:55:33.930]Most people don't get excited about it.
[00:55:35.430]They wanna hear about the science,
[00:55:37.200]they wanna hear about the cool breakthrough, right?
[00:55:39.540]They wanna hear about, man, I found this gene,
[00:55:42.900]and if you do this, it's gonna grow corn.
[00:55:45.060]It's like this and it's gonna be great,
[00:55:46.860]and we're gonna feed everybody
[00:55:48.023]and it's gonna be sustainable.
[00:55:48.900]And everybody goes, yay.
[00:55:50.340]You know?
[00:55:52.170]But, so you have to pitch it as,
[00:55:55.020]I understand these are the societal challenges,
[00:55:57.780]these are the big science questions that we wanna address,
[00:56:00.900]and that's super cool.
[00:56:02.160]Everyone gets all excited.
[00:56:03.240]And then you have to say, okay, but if we're gonna do that,
[00:56:07.260]you need to do this,
[00:56:08.340]we need to get good pipelines and good data.
[00:56:11.430]And that's one thing that NIH has recognized.
[00:56:16.500]Now they have a huge budget.
[00:56:17.430]But I mean they did actually recognize early on
[00:56:19.680]that we're gonna have a data problem.
[00:56:23.160]And most biomedical researchers
[00:56:25.200]can't really do data very well,
[00:56:27.180]and so they set up NCBI,
[00:56:30.150]and they set up the National Library for Medicine
[00:56:33.000]and they invested in some of that infrastructure
[00:56:36.120]and they said, "You guys tell us how to do this stuff."
[00:56:39.240]And it is like huge enabler for research
[00:56:43.080]just to be able to go to one of their sites and go,
[00:56:45.150]oh, they have a whole pipeline
[00:56:47.700]someone just give them my data.
[00:56:48.810]That'd be great. That'd be awesome.
[00:56:51.390]So in these conversations,
[00:56:53.430]I think you're right that we need to pitch it
[00:56:55.950]as these are capacity grants,
[00:56:59.880]this is building critical infrastructure that we need,
[00:57:03.480]and that all of your areas
[00:57:06.870]and all your land grants can all share, right?
[00:57:09.210]It doesn't have to be necessarily like just Nebraska,
[00:57:13.770]or just Florida, or just Washington State or whatever,
[00:57:17.430]but picking different areas
[00:57:19.470]that would focus on different things,
[00:57:21.090]and then going here is the pipeline, right?
[00:57:23.760]Here's what you use.
[00:57:25.350]And yeah, likely there's some chargeback for those services,
[00:57:30.870]but, you know, it's subsidized, right?
[00:57:34.175]It's much less than going commercial, right?
[00:57:35.640]And then when you get your data back, you're like,
[00:57:38.340]I know this is good data,
[00:57:40.140]and oh, here they provided me with a description
[00:57:42.960]that I can just stick in my material and methods, yay,
[00:57:45.450]and I don't have to worry about the data quality
[00:57:48.090]and I don't have to worry about the fact
[00:57:49.440]that I had to process it or my student processed it,
[00:57:54.090]and maybe it's great and maybe it's not.
[00:57:57.000]And I don't have the expertise to tell the difference,
[00:57:59.100]so we're just gonna have to run with it, right?
[00:58:02.250]Just all that.
[00:58:03.083]Yeah, yeah, I agree.
[00:58:06.930]Do we have anything online, maybe?
[00:58:10.384]No people.
[00:58:12.270]They're already on beer 30.
[00:58:13.746](Moderator laughs)
[00:58:14.610]Darn it.
[00:58:17.850]Well, I'm happy to answer other questions.
[00:58:19.920]Like, you know where to find me.
[00:58:22.980]Usually running around campus somewhere.
[00:58:26.160]You can look me up in the UNL directory.
[00:58:29.070]So if things come up,
[00:58:31.380]drop me a note and if I know anything about it,
[00:58:34.500]I'll tell you, and if I don't, I'll try to point you
[00:58:36.650]in the right direction.
[00:58:39.060]Yes.
[00:58:40.350]I'll just follow up with one.
[00:58:41.730]So what advice, if you can sum it up,
[00:58:44.670]would you give researchers in an interdisciplinary space,
[00:58:48.570]maybe an agricultural one here?
[00:58:50.460]And we've got this idea and we're in the idea space,
[00:58:53.100]we're working on an idea, and we get to like,
[00:58:56.287]"Oh yeah, but the data?"
[00:58:57.387]The data management plan maybe of,
[00:58:59.610]you know? Uh-huh. Uh-huh.
[00:59:01.170]Any general recommendations,
[00:59:03.390]best practices-
[00:59:05.023]Best practices From your perspective?
[00:59:06.900]You can sum that up.
[00:59:07.920]Yeah, so there's a couple things bundled in there.
[00:59:12.210]One is, you know, if you can get a group of people together
[00:59:17.790]that are kind of cross-disciplinary
[00:59:19.920]and you can communicate with each other
[00:59:23.520]that you come up with a challenge
[00:59:25.650]or something you want to try to solve, research problem,
[00:59:29.040]and you turn around and you go,
[00:59:30.360]oh, the data, yeah, we need a data management plan.
[00:59:33.090]I'm unlike kudos.
[00:59:34.290]Kudos you got to that point
[00:59:35.730]because most interdisciplinary projects
[00:59:39.870]die in the first six months, right?
[00:59:41.730]Because either you can't talk to each other
[00:59:43.890]or you realize this is a way bigger time investment
[00:59:46.650]than I was planning.
[00:59:48.300]And you think, man, it'd be so much easier
[00:59:50.310]just to go back to my office
[00:59:51.390]and do my own research (indistinct).
[00:59:53.460]So, you know, and a lot of times it's a mismatch of,
[00:59:57.390]not that they don't wanna work together,
[00:59:58.800]but it's a mismatch of what they need out of the project
[01:00:02.670]in order to further their own interest, right?
[01:00:04.680]So there's lots of projects where people come to me
[01:00:06.960]and they say, "I have this great data,
[01:00:09.750]and I have this question I wanna answer,
[01:00:11.070]and like, you could definitely help me."
[01:00:12.990]And then I'm like, "Well, yeah, but I need a publication
[01:00:15.840]where I'm not the middle of 20 authors.
[01:00:19.230]So I need to do something innovative with the data
[01:00:21.210]that's gonna require this part."
[01:00:23.100]And then they're like,
[01:00:24.697]"I just need someone to analyze my data,"
[01:00:26.370]and I'm like, "Okay, then, respect, cool,
[01:00:29.490]I will send you to the statistics core.
[01:00:31.890]And you can ask them, right?
[01:00:32.723]'Cause that's their job, right?"
[01:00:34.080]So sometimes it's just a mismatch
[01:00:36.000]of what people need out of the project.
[01:00:38.580]When it comes to a data management plan,
[01:00:43.080]we do have data management plans available.
[01:00:45.330]Some are easy and some are hard
[01:00:47.970]because for some types of data,
[01:00:50.670]there's a clear place to put it.
[01:00:52.860]It's like this is small data,
[01:00:56.142]it's de-identified.
[01:00:57.750]There's no privacy concerns.
[01:00:59.370]I can go to the library and get a DOI.
[01:01:01.440]Easy, right? Simple.
[01:01:03.600]Other ones, it's like,
[01:01:05.700]okay, we're gonna have five terabytes of data.
[01:01:09.180]What do we do with it?
[01:01:10.650]And the question of federal agencies expect me to keep it
[01:01:14.490]for x number of years after my project.
[01:01:16.590]Who pays for that?
[01:01:18.900]And for a long time you couldn't charge that to grants,
[01:01:22.050]they just said, "No, that's after the grant period.
[01:01:24.600]Like, good luck."
[01:01:26.070]And for sometimes there's a place to put it
[01:01:28.770]and sometimes there's not.
[01:01:30.000]So I think when you're looking at the data plan,
[01:01:32.340]it's like you gotta be thinking about what is the use of it?
[01:01:35.790]How are you going to share it?
[01:01:37.560]Can you put it in a general place
[01:01:40.680]where everyone can access it?
[01:01:43.770]The university does have some systems for that.
[01:01:47.169]We have a system called Research Space
[01:01:48.930]where you can access data that's in different places.
[01:01:52.290]You can even put it in Amazon Cloud,
[01:01:54.600]and it's protected, and it's valid.
[01:01:58.050]It's good for PHI data.
[01:01:59.340]So if there's data you can't let out to the public,
[01:02:02.486]you can put it there, and then you can share.
[01:02:05.400]So the question is kind of like how are you gonna work
[01:02:08.670]with the data?
[01:02:09.600]And then making sure you keep track of it,
[01:02:13.539]'cause the worst thing is, you know,
[01:02:15.630]somebody takes the data and then the student works on it
[01:02:18.420]and does all this cool stuff,
[01:02:19.560]and then no one knows how they did it
[01:02:20.910]and then they graduate
[01:02:21.930]and no one can figure out why we have 11 versions
[01:02:24.300]of the script for this data
[01:02:25.500]because I can't tell the difference
[01:02:26.610]between them and the person's not around anymore.
[01:02:29.610]So using electronic notebook to keep track of things,
[01:02:33.360]using GitHub to keep track of scripts
[01:02:36.810]and processes is super helpful.
[01:02:40.980]And sometimes it's just you learning to listen
[01:02:44.550]to what other person's trying to say,
[01:02:46.830]because, I mean, I know I do it too in statistics,
[01:02:50.130]I use language in professional context
[01:02:53.280]that other people are like,
[01:02:55.334]"I don't know what that word even means.
[01:02:57.030]What are you talking about?"
[01:02:58.230]And I'm like, "Oh yeah."
[01:03:00.300]And I have the same experience, right?
[01:03:02.520]I mean, sometimes I talk to biochemists,
[01:03:04.530]I'm just like, "What are you talking about?
[01:03:07.500]Can you start over?
[01:03:08.670]What the hell is the cell cycle?
[01:03:10.560]Why are we talking about...
[01:03:13.638]We won and the cells...
[01:03:15.458]I am so lost."
[01:03:16.291]So you have to be kind of willing to ask questions
[01:03:18.720]and realize it takes some time
[01:03:22.500]to build like a shared language,
[01:03:25.950]especially across disciplines, yeah.
[01:03:29.790]Excellent. Okay.
[01:03:30.630]Well, thank you.
[01:03:31.500]Everyone, please join me in- Thank you.
[01:03:33.690]Thanking Jennifer, for a great-
[01:03:35.790]Thank you very much. Presentation today.
[01:03:41.400]Yeah, thank you.
[01:03:42.602]Thanks for the invite. Thank you for coming.
[01:03:43.435]Yeah, and please come back.
[01:03:44.550]Join us next week.
[01:03:46.290]Jenny Reese, lessons from extension space.
[01:03:52.138](audience applauds)

The screen size you are trying to search captions on is too small!

You can always jump over to MediaHub and check it out there.

Comments

0 Comments

National and International Agricultural Genome-to-Phenome

Description

Searchable Transcript

Comments icon comment

Related Channels

Comments