Мероприятия Добавить мероприятие Спикеры Доклады Коллекции
 
Продолжительность 47:56
16+
Видео

Lightning Talks Session #1

Shivani Sharma
Nanotechnologist & Cancer Researcher в University of California, Los Angeles
+ 3 докладчика
  • Видео
  • Тезисы
  • Видео
BioC2020
31 июля 2020, Онлайн, USA
BioC2020
Запросить Q&A
BioC2020
Из видеозаписей конференции
BioC2020
Запросить Q&A
Видеозапись
Lightning Talks Session #1
Доступно
В корзине
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
В избранное
69
Мне понравилось 0
Мне не понравилось 0
Доступно
В корзине
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
  • Описание
  • Расшифровка
  • Обсуждение

О докладе

Lightning Talks - Session #1

1:00 PM - 1:55 PM EDT on Friday, 31 July

TALK

Shivani Sharma: Reproducibility in Computational Phylogenetics

Matthew Carlucci: scikick - A Tool for Managing Collections of Dependent Notebooks

Robert Castelo: Mind your dependencies

James Ashmore: Daim: a Bioconductor package for analysis, interpretation, and visualisation of DamID-seq experiments

Spencer Nystrom: Building seamless command line wrappers in R

Jenny Drnevich: Building a Bioinformatics "Middle Class"

Olga Demler: WiNN: Drift Correction by White Noise Normalization for Metabolomic Studies

Moderator: Aedin Culhane

О спикерах

Shivani Sharma
Nanotechnologist & Cancer Researcher в University of California, Los Angeles
Robert Castelo
Associate Professor в Universitat Pompeu Fabra
James Ashmore
Bioinformatician в The University of Edinburgh
Spencer Nystrom
Assistant Professor в Harvard Medical School

I am a PhD student in Keisuke Kaji's and Simon Tomlinson's research groups at the MRC Centre for Regenerative Medicine. My interests lie in the biology of reprogramming, and developing methods to analyse the reprogramming process.

Перейти в профиль
Поделиться

Welcome to the lightning talks, Matthew, and can you present next please, if you want to the previous session on reproducibility, hopefully, this is relevant to the discussion. This is some of the workflow diagram for an old research project. It is a total mess, it doesn't even do all the announcers for the project and this is just some of the Dallas game. Is that you in the first keynote of the conference, to always look at your data? But it's sharing to good reproducibility practices while doing a lot of work involved.

With the gold never doing this again. I work with all that with collaborators to put together a tool for composing and sharing data exploration workbook. Some experienced users. How to make files and other tools that did a good job. But not everyone is comfortable with making web development and we wanted to make something we wanted to use something a little more powerful than make a tool that's user-friendly. We created a simple and sanitize, too old to allow for configurations based on higher level literate, documents without writing custom code.

So using something like this leaves the user with a workflow that can be easily understood by all collaborators. And always acting scripts is rmarkdown, allows users to capture outputs and provide comments to the to the readers and use a snake mix. So you get support for containers and distribute. Getting on your HPC cluster or just a basic tracking of completed results. The title is called psychic and bossy, 2029 marks the public option, but project it's under development and should be used with anyone. Currently struggling to manage their analysis

scripts. They do after their made it up high. But I'm very excited about this will make so much of this, as I can. So this is all available on our studio cloud and anyone can access this link at the top first. I'm just going to make an empty directory. Like I'm starting a new project then I will start run a command to initialize project with some data files. That's just kind of check for some software versions and create some beautiful directories. You can say that here that I have some some notebooks

Next time I'll add these to the North Pole with us. So I have the index page which acts as the homepage of the website and then the key, see a page will import the data from the generates that. So I add the generate stuff as a dependency with the test. This is all started in a simple Jama file, which can be tracked by get the projects designed to work. Well, with Chad and all projects out for good practices. Like not committing large data to get or the results. As I'm just having one more file here to the workflow. Now, running the status command will show.

What is to be executed? Do we have for notebooks here that are all specs? And I'll run status with Josh D. What shows the entire workflow definition. So you can see the dependencies do respond earlier. Now running the workflow with SK. Run starts Nick, Nick in the background and excuse all the chunks. Annual credit report and the report directory. I'm just going to open the South. Are they are marked out on their website which should be smell your and it's based on package down as well as you can see here and automatically created menus

for the notebooks and named the project after the project directory. That's all I can do right now and courage you to check it out. On the rstudio, cloud instance, and see if it would be good for your projects. I would just like to acknowledge tata's careless and Pablo's who contributed to thank you for listening and for this chance to share this work, Thank you so much, Matthew. I'm the next presenter is. Robert Costello who we talked about mind your dependencies.

So this talk is about a discussion we had to develop our solemn. Can you have the wedding to the YouTube video of that discussion? We had about a package dependencies and what you all know that to integrate. August nationality. That's a great thing. But I lost the odds of breaking the package to a chance upstream and you can get a hint about that in the back on Dr. Landing pages for packages where you have a dispatch with the number for Tendencies imported by by

the packets. And and you have a red-colored and this is Pat with cases under 67 in danger, I was something less is more so you may want to reduce the number of appendices. You may have unnecessary dependencies or you may want to dispute the functionality of your package, in two other packages that have a smaller number of dependents East. And how do you do that? Well, that's What we were discussing that we at a couple of functionality for that has a package by a c packets tools and said I'm showing you something like that. That's

already part of the tackett's when you called back asleep and isn't that the frame? And you tell that it is that you were considering the type of the appendices in this case the fence and imports we have stopped Independence. Is that that are going to be installed. When you saw the package, you had a date, the frame with his record, was the name of the packets team, the dependency of the package, that was abandoned, in this case is as important as you see that there are a hundred. Three thousand are also hundreds of thousands of this kind and to get an insight into the message, you can call

the functional. Dependency I craft. You get an object. And in this case I'm going to take a couple packets of such as DC too and just get the graph of the, all the defenses of this package, which is for my 79. If I look at the guitar in the first degree, rape fantasies that said that there was a tarded, the developer has a scratch out, of course you're there. You may have dependencies of other packages of Jeezy black to may have our number of dependents that I'm not working here. I mean, we don't make sense because this could be dangerous if a graph through

cut it. And that's why we have this other functions to help us to get insight into their defenses of our package. So, we stuck at metrics will going to get these four columns, were we have the number of important personalities of that, we get for every dependency than the number of exported functions of a package. And then this is the percentage in G+ to wear in music, to is using blank to find 17% of the functionality of this package. And then the dependency other life is the other life with Dependency dependency, but I thought your package and then this last column is the

dependency game if it's true that. So, if I get through this. I would spare twenty three dependencies on my total number of dependents. You can actually in get more Inside by looking at combinations of of the appendices they are explicit in this case. If I screwed in children and a good game to 11 dependency, but collectively I would spark 15 dependencies. And if I say what, what is the definition of basic to use? If I can get there, they're functional

has been imported by DC too and then I can find out that artistic skulls to ggplot2 that I may consider, whether I could replace them and then go save this number of dependents. He's getting a string by by the music to ABC in Casa Grande. But you said your conference call and talk I was was very inspiring for us to get this done. Marching order Morgan, Charles Johnson, who contributed function package coming in to see Kane and all the developers, who talked about that. If I conduct a package for

analysis interpretation and visualization of Dia on the experiments bed and Brian. For researcher at the center of medicine, I just want you to show you this so you can sing a song similar to check. See where you can at measured dna-protein interactions. So in an example, let's go faster. If we Black Ops 4 wheeler, riding protein is too cold, down in the south. Then wherever you're around the bottom side Digest. Because time itself is a binding protein, you always have to have a control sample. And in this case, the control sample is just down on its own and

dumbass The analysis falls down to a simple. Differential, Bonnie analysis. So for every gatc, fracking in the genome, you want to see which fragments have a higher on this case over the phone. So, as well as the info for the controls, when we see that, we reach out to come to Nexus, move into the genome, which is similar to a taxi and passing sample of the damn only sample. It'll quantify, and the escalation won't eat with your left hand side. If you see, how do you need to do, Samsung Richmond. How big is the difference between morphine and other

faults and we can also write out a quick file so you can look on the genome browser and look on the beat coverage on the corner. There is a genome browser Snapchat. MJC bleeding that they supposed to have to be normalized quite similar to Hi-C. So there's a lot of State personalized. We can bring the coffee options, which will call Yvette on its own. Some to talk Nelly transcription but yesterday she is. Something to eat so much. The next presenter is Spencer Nystrom and he'll be talking about building seamless. Come online rappers in. Our

talk about server side project, I went up working on two Make some analyses, a little easier and smoother. So please So, the premise of this idea is that sometimes, although we might not like it are isn't always the best tool to use in an analysis. So often times, there are other utilities that people have written that work really well and the interface to them is on the command line. And so, that can cause friction, when doing an analysis, where, you know, you're in our, you're in your preferred analysis environment. We have to

exploit the data and run the tool in some way through like a notebook, or some kind of the interface and then import the data back. And so, what I wanted to do was to sort of help. Smooth this process out and make a way to better interface. The two things while using our syntax And so, just a frame of problem, a little more specifically there. There's a few things that are difficult about specifically, like calling shell commands from our and one of those, but I won't talk to you much about his reasoning about

file past, its utilities can be on intuitive depending on how the uses environment is set up. But then the big one is that really passing like multiple flags and user Supply variables is difficult. In the reason it's difficult is because at least to make a call to system, to everything has to be passing in the start of a vector, said, separated Vector, right? Where you pass a flag. But the value for that flag is like a separate entry in the vector and doing computation on that vector and programming with

it. It is just difficult and when I looked around at how other people have solved, both of these problems, There wasn't a systematic way of doing that. And so, that's why I built a package that call command van for taking command function, command arguments and turning them into our functions. And this is just a way to like systematized, the process of actually wrapping shell commands to detail about all the features. But the idea is that it really does more lazy evaluation

of a function arguments and wait, until the shell process is completed and has a bunch of rappers for validating expectation. For basically stating your expectations and checking with a. Those are true for a successful Ron and reasoning about file pads. And so X5 So just a brief overview of how this works. So on the left ear, is an example of a command line tool, which let's say it has to mandatory arguments and be, and then a series of optional arguments one option one. So using

command fun, you can write in our function where you can explicitly defined these as command line functions. And then using the sort of flags junk up here will capture those functions into a vector that you can do. Complication of data structure, you can compute on and then as a final step can crystallize those into a vector, that's amenable to passing the system. And so you can do this lazy evaluation of optional arguments by passing these things to the dots and

with that, I think that's it. I'm so You can check it out on GitHub. There's all kinds of different functionality about it. They see, I still a little fluid, but if people think this is useful and please give feedback on Playstation, can't get Hub and without that's fine, but thanks. Thank you so much. The next presenter is Jenny, who will talk about? Building a bio to Maddox middle class. Thank you. Jenny thinks I'm Jenny drive and I work for the Roy J. Carver biotechnology Center at the University. Illinois, we are core service centers. We have sequencing party on exit

Abalone makes it will have a bye informatics porker that I'm part of us about building a bio for Medics and medic. Middle-class came from an idea from Attack by sea Titus Brown. I saw it about two thousand fifteen on. This is the link to his blog post on an idea. These are my linked to my slides. I also put them in the chat. I've got a lot of embedded links in my size back, Cid, in the middle class, at the end, but I've been heavily involved in teaching, both software data, carpentry, Workshop University of Illinois, for the last five years, the Carpenters have been mentioned several times at

this conference, but for those of you who do not know about them, their target are complete. Beginners and our aim is to teach foundational coding and data science skills to researchers worldwide conference. Our workshops here or at the top of the pyramid or more towards the middle of the pyramid. I myself also teach higher and workshops but all of these really rely on having these grades foundational Concepts and the Carpenters project has really put in a lot of work and studying pedagogy and good research methods about how to teach these people. So, Neal Davis, who's in

population, size in engineering, and all the sides with grass or his, he should have been the driving force behind the Carpenters project at my University would be that every incoming graduate student, no matter what are either in would take a few days carpentry Workshop by the end of the first year. So there's a tree which was first came out and then later that spun out the data, carpentry and more recently, there's a library, carpentry. I don't know that much about the library, but that software carpentry is focused on teaching press. Computing skills regardless of your

discipline in a two-day workshop, with their took before modules will Todd, we use Bosch forecast automation, we will use. Typically, we do get for birth control and then we'll teach one or two modules on programming and scripting. We teach python more than we teach our, but we do our Fair bit and then also we do some Matlab because we have an engineer heavy engineering. Focus O'Neill is also certified to train instructors. And so we have about 20 certified instructors at Illinois to share the volunteer teaching duties cuz we often like to do about 8 to 10 of these supper, carpentry work

shots per year. No date of carpentry covers, many of the same skills but that was tailored to the needs of the particular field on there. Many different lessons and already developed such like ecology and genomics and I've been involved with teaching, but surprisingly there's really not any bioconductor in the genomics lesson and rock. It has been spearheading an effort to create one or more bioconductor related lessons. I took my name is Google group talks. I was not able to go to the first meeting but I promise you, the route that I am interested in working. On this note to get back

to the idea about the buy a maximum class. There's been a recurring theme at this conference about how to help people cross from being a user to being a developer and also do excellent. Birds of feather conference, yesterday on the difficulties of being primarily about a magician. I think these are extremely important and I do not want to talk to them at all, but I also want to point out that it's okay to only be a user. I myself, I am. When I called, how are user or an applied by from Edition. I first started in July, Migrate data 2003, I came to my first by a sea Conference of 2006.

I've been to most of them and I've never developed package and this is the first time I've ever given a talk. Now, Titans Browns idea of the barge Maddox middle-class are those regular researchers who their primary interest is not by emphatic. That so-called wetlab I'll just or medical field your vet students but they do have enough competition skills to be able to use by a conductor packages are other software to do their own data analyses. They don't necessarily need to understand the algorithms behind their functions at this, is not where their strengths lie, but they can contribute by

interacting with developers show. Use cases, file bug report Crestview functionality etcetera. So, conclusion as a community, we need to develop more on ramps, for users. We need to build the base of this big men. Turned it as big as possible because from those users, then you will get more people crossing over to developers from between a strategies. And then also, we need to be supporting the 3 support for navigating the right of your pass. That can be there for a person who was primarily a data scientist. Thank you, thank you so much very important.

Ship, making it available to everybody and buy white noise normalization for Mattel. Potomac studies. Okay? So yeah, so just to give you some background, we have a sew-in but I don't I just show how to report sample samples, what you do, 96 / 384, well, please and then experimental sample. Analyze one by one by the instrument so that is with natural injection order ESO. So we can order a Coke or data from 1 to 2,000, 2,000 samples, and I will cook It's in this is called injection order or an order.

And so then instrument output measure intensity, and then you can download it. And then this is what they expect to see. You expect to see no relation between reality, you see it. But she said here and then drift instrumental drift build up in the instrument and then you shall know. Just, just let me tell you that. This is the one, two, three, four, five regions separated, by physical means a place. So, we have 25 place in, this is play some things 96 measurements. So between plate, nine and Ten's ignition was adjusting the instrument. That's why this

just showing here. So if you would like to How old is drift? And in fact, there is a paper by Jamie Bachelor. Listen to systematically looks at the drift, send me all of your sister. Got your booty. So this is a stable measurements. Then in the middle is the closest aligned by trippie Redd. And then, there is something wrong with it. So the question is, they stole 800 me to buy lights in bed. They classify by type of beat, so they read the most recent in the snow

and then a brown and being. So, each and light intensity Sportage instrument. And let me slide so I can buy play. So if you don't, do you spell component analysis like here in fluid components or component, cords is not a component for your data and color it with the other. Please, write to each daughter is a is a, is a well, and so we call it ignored you, this plate number. And then, so, there should be no lasting by place. But he have you seen rolled a huge

Blessing by plate when we applied but it is still present to hear by afternoon Miss that is passing by point five to rlcs. So therefore he needs to develop our own Signature Collection method. And the assumption that I was signal You should not be related and independent from the running order. So, essentially it should be white noise. So I can I cast from analyzing usual and solenoids is that, we'll just annoys you want and then why it is? Because you can see this

right on top then we each plate vehicle on 3rd and Division, and Medium across play. This is your step. And then meet us to White Noise because we don't want to over, correct? So you want to be gentle as possible to our location to be as gentle as possible. So he's okay. So I'll show you what it looks like. So, this is their connection. This is Sensational stories with Distortion. And then black line is what's a metrical calculate. Then she is a correction threshold of

20%. So 70 for metabolites before might have to be applied on a 2025 with a stressful and then we also did some more Corrections and guess what do this before? And after looks like Yeah, and this is how I almost fell asleep. Thank you for your attention. Thank you so much. Thank you so much. I know what the lightning till format is quite fast. The next talk is and some Shabana, Shabana Sharman and it's reproducibility in computational phylogenetics 2 challenges

in experimental studies of experimental conditions that may be attributed to a lack of experience and reproducibility. He's definitely accepted all adapted by others of the same Community. Action is possible in case of very interested, and some type of this is gets just moved on Long jumps. Listen analysis, showed that can you turn on the next flight? Recent analysis. Showed that approximately 25% of the studies need the sequence alignment, Sentry's abilities. And additional

data was used for only 42% of the studies of any kind, not onset of several Community initiatives to improve reservation and sharing a phylogenetic data, has helped the disposition of qualitative data to online archives. The best part of reproducibility of an experiment and results that are the same everywhere. Next. Next flight. Adaptation of an evolutionary tree. Experiments in a video that someone outside their team is able to do the experiment and get the same tree.

Next. A plethora of tools are available for implementing reproducibility such as Galaxy platform. That allows one to construct an individual's parameters and databases such as internal for fetching analyzing and processing. The data station energy produced by genetic such as Pelosi. Under single utilisation for static analysis, biogenetics. But nobody seems to be losing them, so I think in this 3 minutes. Thank you so much, but was a wonderful world introduction to Manny. Feels that are important

for by a conductor That's actually go to question for mass and I'm going to just share the screen. The questions, please put your questions in the pool and rank of questions that you like for the speakers. Please ask the speakers so that we know which questions speakers van. So, the first question and it's got the highest ranking is on Spencer and thank you for the command function. How do you use process X and what is the main difference between your package? And there's a

sort of where my package comes to? My package is completely compatible with process tax and what it is doing is its systematizing, the everything just Before you run the command. So getting it all ready to send to either system to or process tax. In fact, whenever I use this package in my own development I use process tax because it's process handling is much better. So where my package comes in and is dealing with Getting to the Ron and Dan after the run. Expecting that the outputs are correct and even doing things like parsing the help. You know, if the if you get an

error it does things like parse the help file of the tool, you are trying to use and suggest reasons why that might fail. Thank you, Robert. You need only one function from package. A you dropped that code import packaging, copy paste, the code into the package and site Packaging. Not necessarily I mean if you are importing a package for one function, but that's functional feeling really crucial for what your packet is doing. You shouldn't drop that the dependency show

me, they need more to identify dependencies. That are not that important centers in their life, cycle of a package, things change through time. And then you realize that something you would import naturally you're not using that much anymore, or you have protected the couch, and I see the elf of your that, you have always seen that. You only put a function, you look at that. And then you decide whether that's really important or not. Not how does your workflow with our scripts and our Command? And how do you determine a bond order

functions call Star scripts automatically as a notebooks, using their spin? So running our script. Who provide you all the same output that's something that are marked down. I swear that they're on order, the order is a specified through through the dependency needs to run before another, the statement back and will choose an appropriate, run order for those tasks and you can paralyze them from the chat Simone as well as the poles. Let me check this. Thank you.

Jenny. I had some questions for you regarding the cop and trees and training and how we might best sharp provide training to buy conductor users in the use of survey. Last year training was identified and particular instructor training material was identified as a great need within the community. Given your experience within the cockatrice. How do you think we can improve? Well it's it's interesting so teaching bioconductor as an introductory Workshop, I struggled this, what what does it mean to be

bioconductor? And so there's you know, you know, you know, your base, our stuff in the coding and that's kind of a separate introductory workshop and then bioconductor I find it's more that I when I teach it I shall keep here's the community. Here's how you navigate the packages. Here's how you find this or that and then I often stress that cuz if you're coming from and you're coming in, I learned our because I needed to do data analyses. Somebody had already written the code, I don't need to write the code. I need to be able to follow their code from their vignette or work slow and

nowhere to change things. So that's kind of what I stress in my workshop is. Oh look, here's an example code here. You know, here's how you extract, all the vignettes for this package. Now, I want you to do it to this other package. And just kind of that way. And I don't know if that's the best way or not because they're so many packages. I do also go over the s4optik structure cuz that is a big difference from bioconductor than anyone else right now though this the next year. So, it's going to be interesting with trying to teach remotely is what

we're looking at. Recorded workshops are teaching, remotely is not nearly as good as having somebody on site, being able to help you over your shoulder. So who knows how that's going to be a? I had to really try to think of how do we make like a one-hour Workshop interactive and I found a really, really difficult. Yeah, there's not enough time. Yeah, so sorry. So in the check their comments, but no real questions. So it looks like the speakers also reply to each other. So, there are only you really do

this for questions in the pools. I don't know if they question 2 math has been addressed And the question to Robert's Hair. Robert, you only need one function on a cell phone if you break your package and two smaller ones. Do you venwise workflow package? That shows how old is the work together or is it about describing the smaller packages in the VIN? Yes, so suggest. Well, as an important depends, it seems like you end up with an artificial dependencies and same Upstream. Dependency issues in the end unless you drop the dependencies

across all of your smaller packages. So it sounds like trying to reduce dependencies and what's the best approach to do us? Well, when you try to register Tendencies, usually, you should take an approaching with you and personality. The typical situation would be that your package has made some calculations. Very a parcel. A lot of dependency, some of you would like to break their calculations. And then of course a set of the bed is not going to be probably more of the joint between these two packages

cream. It will break maybe through each other or the same time for splitting. The packets dependencies to make it more robust. You don't have to fix the package. One part of the package because something has to have nothing to do broke and India. James. I have a question for you which is I'm actually not that familiar with Gia AMC. Can you actually just describe what what is actually doing a little bit more? So it was originally, it says Sullivan old technology and it's just been using the mailing systems. So in

situation because you don't have enough cells, you can use the Magnificent. What happens if you take protein base to it, express that fusion protein in the cell. And then from there, you can. ATC sites at the mouth and Stevenson and where your protein is found in the female version of Chips baked Most sensitive interesting situation for you count is cheap seats. So if you don't have a chip grade antibody, if you want to do tissue-specific profiling, if the sample with juice plus a $10,000 as opposed to a million cells

is 50 usually has. It's really for those social situations. You mentioned that there is a reproducible phylogenetic workflow, and buy a conductor that can be applicable to plant genomes to plant. Because I did not mention that, okay? And how many species can I cover? Okay, thank you. I'm Jenny is the division of labor principle, not that sure as individual level rather than trying to teach every individual, everything creation unicorns, would it not be better to establish Labs with students from diverse backgrounds and let the plates

there? Interests that should help. I am from editions Publishers first order papers and experimental biology I mean that it is a good strategy. We had there was lots of discussions yesterday at the birds of the feather about the poor, the buy-in for a tax person attached to the experimental lab that they're lonely and stop and don't get the credit for the stuff. I think I would argue that if you might be better off while you have diverse people that they also trained each other. So you have redundancy in your lap. You don't have just one person. I mean, I've had at

least four five situations where he has kind of me. I had a postdoc. You did this at postdoc has not left and we no longer have any that capability in our lap. I'm so you really, I think it's a p. I that you should try to encourage everybody in your lab to get some of the skills. Authorship is a weird weird. Weird is situation. It can be hard. A lot of times you just I actually got one first author as the buying from a Titian but it was a betta survey of like 15 different studies on zebra finches that I did all the analyses for

and then I did The Joint analysis and I wasn't even primary person to write up the paper. But the key item in PE I said wait until Janet you were always V author and all these papers were going to give you first. Saw the ship on the but that's something that was gifted to me. Not that I could like ask for kind of Thank you. I'm sorry. We have 2 minutes left. We will copy the questions into the slack channel. For people that want to continue the conversation Olga you were talking about metabolomic, I think that's

a feels. There's a lot of people are very unfamiliar with us and you talked about very important question of noise in these studies. Do you think this noise metabolomics is worse and more difficult to handle than those in the more more widely used and sequencing now? Well, I think in the Bible makes it is kind of easier because it is not really annoys. Well, that is nice present but actually there are more Drifter noise and research easy to regress. You know, you're out if you have a few annoys

agent to date and then this is a very bad problem. You cannot shoot it out noise, you can expect me to buy Lumix intensities to be annoyed. And now we have added another noise on top of that noise. So this kind of approach. Yeah, but it is a situation when you think your instrument that uses Thrift whether there are some systematic differences I'm going to ask one question of James, Robert, and, and map between a lot of questions for you. And I know that would be something that will continue because

optimizing how you run your packages and so important. James, I want to ask you a question which is from damn. How does your statistical approach compares, the original baskets in the head and cough lab particular, thinking of how they don't normalize the top 10% of the deja since it's the real signal and which always seemed weird to me to be questioned. So different ways to bet so much easier and Pisces you'll have to buy CDs and when you're trying to compare and if they have this effect, Lcctc,

counseling to CBC has a lens and a route to 545 Midland fire station, which I think many people have seen a distribution between the dam and the dumb fusion protein separate. Okay. I have an interesting thing for my two computers are showing me different times, quite time for one more question or we'll relieve the last people are interested in. Let me know a lot of people asking these questions of proper term ass would probably be very interested to see Aaron's talk. So maybe we should

give people a break before they see the Keno's. So I think we allow, thank you for an extraordinary lightning and session. I really enjoy this and I learned lots. And I hope you did too. Thank you so much.

Купить этот доклад

Доступ к видеозаписи доклада «Lightning Talks Session #1»
Доступно
В корзине
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно

Ticket

Доступ к записям всех докладов «BioC2020»
Доступно
В корзине
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Билет

Интересуетесь тематикой «Наука и исследования»?

Возможно, вас заинтересуют видеозаписи с этого мероприятия

27-31 июля 2020
Онлайн
45
19,14 K
bioc2020, bioconductor , dna methylation, epidemiology, functional enrichment, human rna, probabilistic gene, public data resources, visualizations

Похожие доклады

Atul Deshpande
Postdoctoral Researcher в The Johns Hopkins University
+ 1 докладчик
Daniel Bunis
Biological Data Scientist and Immunologist в University of California, San Francisco
+ 1 докладчик
Доступно
В корзине
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Доступно
В корзине
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Qian Liu
Assistant Professor в Roswell Park Comprehensive Cancer Center
+ 1 докладчик
Qiang Hu
Postdoc в Roswell Park Comprehensive Cancer Center
+ 1 докладчик
Доступно
В корзине
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Доступно
В корзине
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Julie Zhu
Professor and Head of Bioinformatics Core в University of Massachusetts Medical School
+ 1 докладчик
Jianhong Ou
Bioinformatician в Duke University
+ 1 докладчик
Доступно
В корзине
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Доступно
В корзине
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно

Купить это видео

Видеозапись
Доступ к видеозаписи доклада «Lightning Talks Session #1»
Доступно
В корзине
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно
Бесплатно

Conference Cast

ConferenceCast.tv — архив видеозаписей докладов и конференций.
С этим сервисом вы можете найти интересные лекции специально для вас!

Conference Cast
1497 конференций
47700 докладчиков
20185 часов контента
Shivani Sharma
Robert Castelo
James Ashmore
Spencer Nystrom