Now I'm on sabbatical I finally have time to make more OMGenomics videos! Starting with this one:
Please watch and leave comments about where you are at in your skill-building journey :)
Nice Nature article on visualization in genomics! Some nice discussion from
@asiepel
of disseminating results through creating new browser tracks, and even a few nuggets at the end from yours truly 😅
Does anyone have project ideas for undergrads or early grad students in bioinformatics/genomics? Maybe interesting papers where the data is actually available to reproduce the results?
DeepVariant 1.0 is released!
Notably: "Compared to DeepVariant v0.10, these changes reduce Illumina WGS errors by 24%, exome errors by 19%, and PacBio errors by 52%."
Plus a new hybrid PacBio+Illumina model:
Exciting news!
@RobAboukhalil
and I have just launched The OMGenomics Podcast! 🥳
Featuring casual conversations about the non-scientific parts of life in the biotech industry.
Episode 1: Sabbaticals
➡️Apple Podcasts:
➡️Spotify:
The "DeepVariant 1.0" talk I gave at the
@CSHL
BioData meeting is now on YouTube!
It covers how DeepVariant's pileup images and shows some of the many improvements in accuracy, runtime, and new models the team has made since it was open-sourced in 2017.
Just launched Chordial! Powerful, fast, easy-to-use software for creating chord diagrams. . If you're following along, yes, Chordial is Circa's little sister (and I will love them both equally) :)
Wow I'm a finalist for the The Benjamin Franklin Award for Open Access in the Life Sciences! I guess this is the part where I humbly ask for your vote? 😅
I made a tutorial on *GNU parallel* -- a new video on my OMGenomics channel covering everything you need to know to use GNU parallel painlessly, including an example of using parallel for launching bioinformatics jobs to a cloud or compute cluster.
What are interesting small businesses in the bioinformatics space? Looking for bootstrapped (not VC funded), profitable businesses. We may want to interview them for the OMGenomics podcast 🧐
I used some time over the winter break to make a new video with "5 tips for better bioinformatics software" mostly based on things I have learned to do differently over time (a.k.a. my own mistakes).
I just released a new OMGenomics video:
How to edit files on the command line with vim
I cut the tutorial down to just the most practical features I can’t get by without.
Not gonna lie, I learned several new things myself just by making this video…
Watch
@RobAboukhalil
and me test-drive GitHub copilot for bioinformatics algorithms! The results were impressive and hilarious. Just wait to see how it generates lunch ideas for us at the end 😂 🍎🥕🍇🍅
OMG(enomics 😝)! I just finished porting my website off of WordPress and onto Jekyll+GitHub pages! Oh it feels so good to be able to edit the site just like I'm writing code and committing it to a git repo. 😎
Another OMGenomics podcast episode is out! We interview
@nilshomer
about his experience and wisdom from co-founding Fulcrum Genomics, a bioinformatics consulting business helping biotechs solve interesting and novel problems.
Exciting news to share: I'm building a data visualization team at
@DNAnexus
! Looking for a great data vis software engineer:
#AGBT18
Please retweet if anyone you know might be interested.
Our research using long-read sequencing to find structural variations in breast cancer made the cover of Genome Research! Thanks
@mike_schatz
and everyone else who put countless hours into this project
DNAnexus is looking for summer interns with biology and coding experience:
We have a lot of interesting projects going on, including some in genomics data visualization. Please forward to anyone who might be interested!
Now we not only have a paper but also a fancy article about the paper :) "Massive genome havoc in breast cancer is revealed" via
@genomeresearch
@EurekAlert
THIS is how you make a genomics cloud platform that is actually specific to genomic data.
Disclaimer: I helped with some visualization and UX ideas for this :)
Excited to launch , a free platform for exploring genomics data (anything public on AWS, GCP)
Featuring:
- Interactive previews
- Sub-sampling
- Grouped indexes for .bam/.bai, etc
- IGV integration
- Interpreted CIGAR strings, SAM flags, quality scores
🥳I’m so glad our work overhauling half of DeepConsensus for speed optimizations is ready to release today! It took a lot of work from the whole team, but especially thanks to
@daniel_e_cook
for organizing our efforts!
Release of DeepConsensus v0.2. Large speed improvement (>10x). Added GPU support (w/ NVIDIA V100 is ~3.3x faster than CPU-only). Added Docker images and streamlined dependencies for ease of use. Improved CLI interface and one-step execution script.
()
Check out this new blog post from
@gunjan_baid
about DeepVariant, giving a great overview of all the major changes and improvements the team has made since DV was first open-sourced in 2017.
New OMGenomics video: "Should you do a PhD?" As always this is just my perspective as someone who's a bit further down that path and has had this conversation with friends and colleagues before!
Check out this interview Microsoft did with me for their
#WomenInSTEM
blog! We talk about how I got into this wonderfully interdisciplinary field, and about the work we are doing at
@dnanexus
"After we've put all this effort into building one assembly for a species, shouldn't the next one be easier?" --
@mike_schatz
advocating for a resurgence of reference-guided genome assembly in a great talk today at
#PAGXXVI
A Circa user had a question, so I got carried away and made a whole video to answer them: "Circa for RNAseq: Differential expression circos plot"
That was pretty fun actually, I should make more little technical videos like this.
Release of DeepConsensus v0.3. ~5x faster than v0.2. Improved accuracy (yield of Q30 reads on a 24kb dataset increased to +49% relative to pbccs vs. +41% with v0.2).
()
Amazing tech lead job by
@marianattestad
, with
@pichuan
as Eng manager.
Excited to release a new OMGenomics podcast where we interview
@kajawasik
and dive deep into one of her companies, Variant Bio, whose goal is to develop new therapeutics by studying the genetics of populations with exceptional traits.
New episode of the OMGenomics podcast just dropped! 🥳
This time with video and a guest: our friend Steve Osazuwa!
We answer the eternal "should I do a PhD?" question and hear what Steve's barber thinks.
(Also find it wherever you listen to podcasts)
Fun side effect of doing data visualization in my normal work is that I wasn't satisfied with Mint's charts for analyzing my spending habits. So I did it myself in python and made a video about it:
🚀 New on : solve Rosalind coding challenges with an IDE and a Python interpreter, directly in your browser!
Check it out:
First 10 exercises are up, more to come! cc
@PhillipCompeau
We just released episode 3 of the OMGenomics podcast: "Job hopping".
Inspired by opinions seen on Twitter, we discuss whether people should stay much more than 1 year per company to avoid being judged by future employers as a job-hopper.