Nils Homer
@nilshomer
Followers
5,570
Following
746
Media
1,709
Statuses
7,058
I write software for Biologists. Founder @fulcrumgenomics , Bioinformatician, Computer Scientist, Coder, Father of 2xGirls. Opinions are my own.
Phoenix, AZ
Joined July 2009
Don't wanna be here?
Send us removal request.
Explore trending content on Musk Viewer
Leverkusen
• 421293 Tweets
#Comebacktome
• 267550 Tweets
RM IS COMING
• 209480 Tweets
メイドの日
• 85735 Tweets
Celtics
• 77211 Tweets
Luka
• 69445 Tweets
Maranhão
• 61388 Tweets
WIN CENTRAL FAMILY
• 54610 Tweets
#ดวงใจเทวพรหมSportsDay
• 47589 Tweets
Rangers
• 44119 Tweets
メイドさん
• 40145 Tweets
Cavs
• 37859 Tweets
भगवान परशुराम
• 37670 Tweets
Tatum
• 33394 Tweets
メイド服
• 31495 Tweets
ドラマ化
• 27797 Tweets
Safe Campaign
• 25535 Tweets
भगवान विष्णु
• 22702 Tweets
Canes
• 22600 Tweets
PRISCILLA
• 17931 Tweets
RTTT
• 13805 Tweets
#VileirosNaMansão
• 12652 Tweets
Pumas
• 12443 Tweets
Rex Murphy
• 12403 Tweets
Cruz Azul
• 12333 Tweets
Zavala
• 11690 Tweets
Aleska
• 10859 Tweets
Pinned Tweet
Join us
@fulcrumgenomics
!
Nowhere else can you have access to so many diverse and challenging
#Bioinformatics
and
#Engineering
problems across so many clients and domains:
‣ 10+ 5 year clients
‣ 100+ projects completed
‣ 50+ clients served
We're hiring
@fulcrumgenomics
! We're looking for an experienced bioinformatician to join our team and work on exciting projects in technology development and biomedical research. If you're interested in becoming part of our team, drop us a line here:
4
16
32
2
28
79
My children are the world's most successful people, followed by their parents 15 minutes later.
11
117
2K
Experienced Bioinformaticians:
Comment below the one thing you'd like to tell junior Bioinformaticians.
New Bioinformaticians:
Read below for an awesome list of career tips.
198
530
2K
90% of
#Bioinformatics
is knowing a tool exists.
The rest is knowing all the tricks to get it installed.
33
154
1K
In response to this unoriginal tweet, I’ll go first:
Never trust your first results, especially when they agree with what you expected. Skepticism is the price we pay for amazing science.
6
20
449
I am proud to announce the launch of ExcelBio!
Driven by overwhelming demand from Biologists: build, manage, and execute your
#Bioinformatic
workflows all from your favorite GUI, Excel. We are proud to support Illumina SampleSheets as our first product. DM me for a demo!
20
26
273
#Bioinformatics
software lifecycle:
Someone has written this before.
All of them are unsupported, and I don’t want to pay for it.
Fine, I’ll write my own, I can do a better job anyhow.
Job’s done, but no money for maintenance or support.
We now have N+1 BFX software.
10
35
219
#Bioinformatics
on AWS has never been easier!
So long CloudFormation; say hello to the
#AWS
Genomics CLI. I've used it with snakemake, nextflow, cromwell, and miniwdl. It takes the pain of going from a working workflow locally to one on the cloud: 🧑🍳💋.
6
50
223
🎉🎉New samtools release!🎉🎉
I’m really excited about this release for one reason: template coordinate sorting. Let me tell you why…
Release 1.16 of HTSlib, SAMtools and BCFtools is now available from GitHub. See for links to tarballs and release notes.
0
39
108
2
30
189
I have a confession to make...
I have replaced writing bash scripts with
#Snakemake
for all my
#Bioinformatics
analyses, even trivial ones...
I'll let you figure out why.
26
13
183
Search a FASTQ for specific ref & alt alleles with fqgrep
Yet another little
#Rust
tool for
#Bioinformatics
; install with
#bioconda
; developed at
@fulcrumgenomics
.
See more:
2
30
179
Surgery complete… looking great, feeling awful. F*ck cancer.
17
1
170
Years of my life could have been saved if Nature papers had this in their supplementary material…
Excited to share a new preprint with
@XiChenUoM
and
@lpachter
introducing seqspec for describing, organizing, and annotating elements in sequencing libraries. seqspec is a machine readable specification + an associated suite of tools. 1/9
8
78
326
4
14
143
@sebatlab
Great folks coming out of undergrad are getting six figure
#Bioinformatics
jobs, let alone out of a PhD or post doc. They both need A TON of training, but they’re worth it. Whether or not you like it, you’re competing with us.
5
5
131
Breaking: CDC says fully-vaccinated people can go back to using TopHat again, with all custom scripts written in PERL and committed to CVS (not the pharmacy).
#Bioinformatics
7
11
130
By the age of 30, you should have uninstalled/reinstalled conda 5 times with over 100 PERL scripts saved that you don't have any plan to actually use.
#Bioinformatics
13
8
136
Dear academics, I'm going to eventually read your code. Please feel the appropriate amount of shame from the start about your hard-coded paths, missing required files, and major assumptions about file naming. Your frenemy, industry.
7
9
136
I love writing
#Bioinformatics
tools, plumbing them into pipelines, written with care, and seeing them produce novel Biological insights at scale.
If that’s you, come join us
@fulcrumgenomics
1
16
133
Do you know all these
#Bioinformatic
tools?
𝚝𝚘𝚘𝚕𝚔𝚒𝚝 | 𝚝𝚘𝚘𝚕𝚜
----------- | -----
𝚋𝚠𝚊 | 𝟷𝟺
𝚋𝚌𝚏𝚝𝚘𝚘𝚕𝚜 | 𝟸𝟷
𝚜𝚊𝚖𝚝𝚘𝚘𝚕𝚜 | 𝟹𝟹
𝚋𝚎𝚍𝚝𝚘𝚘𝚕𝚜 | 𝟺0
𝚏𝚐𝚋𝚒𝚘 | 𝟻𝟸
𝚙𝚒𝚌𝚊𝚛𝚍 | 𝟷𝟷𝟿
𝚐𝚊𝚝𝚔𝟺 | 𝟸𝟽𝟻
18
7
127
Sometimes I wonder if people even use Picard tools in
#Bioinformatics
Time to get a bit personal about how most of my tools “never made it”, imposter syndrome, and it’s ok to fail.
A thread 1/n
22
18
127
How do I keep my scripts and data perfectly organized in a 40 hour work week while raising two young kids?
#Bioinformatics
#DataScience
🧵
9
20
122
It’s just lab work. Can’t you just sequence each chromosomes individually? Data ready tomorrow right?
Translating how some Lab Scientists talk about
#Bioinformatics
. What’s your experience?
14
20
105
Captain’s Log, Supplemental: pathology report clean, great margins, no evidence of malignancy. Feeling great in addition to looking great.
Surgery complete… looking great, feeling awful. F*ck cancer.
17
1
170
3
2
112
“Effectively cured” wasn’t a phrase I was expecting in my timeline, but I’m incredibly grateful.
5
1
112
Some of the things I’ve worked over the past year:
> cancer vaccines
> genome editing
> novel DNA sequencers
> single cell-finding algos
> clinical diagnostics
> forensics genealogy
> modified oligonucleotides
> open source software
> high performance file systems
> file formats
5
6
96
We’re seeing organic adoption of
@LatchBio
by more than a few of our clients. Here’s why we are very positive on the platform (hint: Biologists running Snakemake), our excellent experiences working with the Latch team despite past public behavior, and overall thoughts. Thread…
8
17
93
Four serious DMs so far. I am so sorry.
I am proud to announce the launch of ExcelBio!
Driven by overwhelming demand from Biologists: build, manage, and execute your
#Bioinformatic
workflows all from your favorite GUI, Excel. We are proud to support Illumina SampleSheets as our first product. DM me for a demo!
20
26
273
9
2
91
See the slides, you must really see the slides:
4
21
91
Trying to recruit the best Bioinformaticians over here.
Fair and equitable pay, great life-work balance, engaging and challenging problems, world class clients and
#Bioinformatics
team.
I have only but one problem to solve…
16
14
93
fqtk 0.3.0 released with fixes and improvements!
Demultiplex your FASTQs in
#rust
!
Also available in
#bioconda
and
#Nextflow
(via
@nf_core
/demultiplex). See more at: .
#Bioinformatics
3
21
93
grep, but for FASTQS, but now more grep-like
We've done a lot of work
@fulcrumgenomics
to try emulate grep, but for FASTQs. Almost all of the grep command line arguments are now implemented.
Try it out and submit your bug reports.
#Bioinformatics
grep, but for FASTQs
I’m looking for folks to try out a major upgrade to fqgrep that makes the usage almost the same as vanilla grep.
Please build off of this branch while I add unit tests:
2
5
50
1
17
88
Now sequencing vendors are not providing raw FASTQs (available upon request, denied). Did they learn from academia?
20
9
84
All genomic assays papers should have an IGV screenshot to show me how the reads will look after all that effort.
4
1
80
We in
#Bioinformatics
need to write unit tests.
I almost always ready tests first. They are the best documentation for how the author intended for the code to be used and behave.
Write at least a few to help me on my way.
Tip: Starting code reviews by looking at tests is a good way to get a sense of what the change is about (it'll reveal things often hard to spot in the change itself). If there are no tests, stop reviewing and ask for tests.
28
164
1K
13
12
78
A billion here, a billion there, and the rest of us just trying to support open source software on which genomics stands...
Exciting news from
@Verilylifesci
on this $1Bn funding to further our precision health impact, including evidence generation. Big congrats to
@stephengillett
, who will be our new CEO, and deep gratitude to Andy Conrad. We're set up for great things ahead!
1
18
74
7
6
82
Building a
#BioInformatics
tool repository to me feels like the same as annotation and curation for variant databases:
1. garbage in, garbage out
2. everything is done manually
3. relying on academic grade tools (data sources) is fraught with peril
We just raised a $28 million Series A from Coatue and
@Lux_Capital
to build the AWS x GitHub for biocomputing.
1/N🔻
17
83
724
10
9
79
FOMOmics: the field where only the newest and coolest genomic technologies are bought and applied.
3
10
80
I'm seeing wide-spread layoffs in
#Bioinformatics
and
#biotech
.
So many good and talented people affected; I'm here to listen and help in my own small way.
6
11
77
For all the excitement of
#Bioinformatics
software in new programming languages, the API designs and the ease of use for other developers are… not good. That’s why I sometimes call it academic grade. Gets the paper published, but helps no one later.
3
10
71
1
27
74
Come work with us
@fulcrumgenomics
!
Bioinformaticians and Computational Biologists welcome!
We solve interesting biological problems through engineering and bioinformatics for a diverse set of clients. We also value life outside of work. See more below.
#Bioinformatics
1/n
2
38
68
“Validated in IGV”
My trust level: 📈
Can we normalize adding IGV screenshots to a supplementary PDF?
7
3
70
I don’t work with bullies.
I was bullied as a kid, and now I have the privilege to choose with whom I work. I don’t need you, you need me. Thank you next.
2
7
65
Why don’t journals have a read only GitHub organization that authors must deposit code (and small datasets) into? I’ve been burned by folks moving, deleting, or overwriting.
Same for wet-lab protocols not described in the paper or supplementary materials.
9
16
68
My master plan
1. Joint appointments at Sanger, EBI, and Broad
2. Standardize-h/-H in samtools and bcftools
3. Quit
My master plan
1. Get hired by illumina
2. Release the 2x400bp kits they developed ten years ago
3. Quit
7
7
192
6
2
66
It’s 2022 and I’m still writing aligners!
#Bioinformatics
~13 years ago I was asked why I was writing an "aligner" as the problem was solved. Today, I'm writing a new "aligner".
#ThisIsBioinformatics
1
3
37
3
10
70
95% of data science is cleaning the data
My students: So when are we going to learn machine learning models?
Me:
7
6
54
4
12
66
#Rust
is quietly becoming our go to
#Bioinformatics
performance-focused language. Check out some of the libraries we’ve been working on.
2
15
43
0
8
62
Bioinformaticians: which one(s) do you consider yourself?
• Molecular Biology Debugger
• Technical Leader
• Thought Leader
• Expert Executor
• Pragmatic Skeptic
• Trusted Teammate
#Bioinformatics
#Hiring
#Genomics
@fulcrumgenomics
14
14
65
#Bioinformatics
is a lot about persistence.
It’s not about how many times you get it wrong, it’s about if you persist long enough to get it right.
9
6
62
I want IGV for complex rearrangements, to visualize long reads aligned in graph (not a dag) form against many short contigs (non-linear reference) or distant segments (linear reference). Imagine a SAM file with many complex non-linear chimeric alignments. This must exist.
11
7
63
Who says you can’t do open source
#Bioinformatics
in your day job?
NB: we’re hiring.
#fgbio
2.0, the second major release, is now available.
A major theme of this release is performance of the UMI-related tools. And for those using picard's MergeBamAlignment, try using fgbio ZipperBams instead. Same author, new tool.
#Bioinformatics
0
12
45
2
9
60
I am an author on the original SAM specification publication. It's not just you.
#Bioinformatics
2
3
60
There are so many intangible benefits supporting users of your open source
#Bioinformatics
, but at some point you have to get paid for it. The folks asking the questions are getting paid, so why not us?
5
5
62
@bernhardsson
Most bioinformatic tools are command line based, have wildly different compute resource requirements, are data/file heavy, and are insidious to install/containerize. Airflow, Luigi, and others are not built with those things in mind. And so we have our own.
3
2
60
At some point R and assembly code are going to be indistinguishable.
2
2
59
SnakeLatch?
I am genuinely excited at the prospect of a
#Bioinformatics
platform that has an amazing UI for Biologists on which Bioinformaticians can develop using Snakemake.
2
10
62
How does one leverage a genetic database of millions of samples to solve cold cases or identify human remains? To find even 4th degree relatives with only 10K SNPs?
#Bioinformatics
#Forensics
A 🧵…
1
14
58
My new aligner is awesome.
It’s not meant to be fast, but instead gets to single molecule base-pair resolution of some very interesting events. I can’t say what yet, since no one is going to see it for a while, but it’s awesome. I just need to tell people.
2
0
59
Alignment is solved they said...
8
25
58
What’s the BWA of CNV calling?
It still feels like a dark art where you need to roll your own.
#Bioinformatics
14
9
57
Maybe making people click ads shouldn’t be paid multiple times more than folks developing the next cancer vaccine or gene editing therapeutic. 😡
Will there ever be good digital infrastructure in biotech? Alternate title: "Part of the reason it's hard to hire developers in bio is because they don't have the tools to succeed that they do elsewhere - will that ever change?"
3
28
136
2
1
57
How to differentiate yourself in
#Bioinformatics
and
#CompBio
:
Learn a “performant” language in addition to “data science” languages.
Former: eg. Rust, C/C++, JVM
Latter: eg. Python, R, PERL
8
7
49
@dgmacarthur
Heng is exceptional at many things, but one thing that is rarely talked about. He skates to where the puck will be. And has a toolkit ready for when the data arrives.
0
5
52
Now do IGV screenshots with real variants versus sequencing artifacts.
3
2
51
Move over samtools, sqlite3 is here to take over.
TIL you can run SQL queries directly against CSV files as a one-liner using the default sqlite3 command line utility
57
779
5K
3
5
47
Join us fully
#remote
@fulcrumgenomics
We're looking for the experts in
#Bioinformatics
.
Requirements:
‣ Curious, Conscientious, Client-Focused
‣ Demonstrated expertise in a
#Bioinformatics
area, familiarity in a few others
‣ US-based
#hiring
#jobs
#genomics
Join us
@fulcrumgenomics
!
Nowhere else can you have access to so many diverse and challenging
#Bioinformatics
and
#Engineering
problems across so many clients and domains:
‣ 10+ 5 year clients
‣ 100+ projects completed
‣ 50+ clients served
2
28
79
3
29
51
#snakemake
users having issues installing today, see this issue and fix (downgrade tabulate):
#bioinformatics
0
13
48
grep, but for FASTQs
I’m looking for folks to try out a major upgrade to fqgrep that makes the usage almost the same as vanilla grep.
Please build off of this branch while I add unit tests:
2
5
50
Academics want me to peer review and don’t want to compensate me for it?
It’s supplanting actual client work where I’m getting paid: your lack of budgeting is not my emergency.
And also, perhaps if you paid reviewers you’d actually get some.
5
1
47
🎵🎵🎵 Don’t be fooled by the socks that I got,
I’m still, I’m still the Bioinformatician on the block.
I used to have short reads, now they’re long.
No matter what I sequence, I know where they map to.🎵🎵🎵
(Thank you
@PacBio
for the socks!)
1
1
49
Want a feature request in a open source
#Bioinformatics
software? Sponsor it!
So many of our
#Bioinformatics
tools and features have come about through industry sponsorship and giving back (we love it).
But…
2
8
49
Bioinformatics is job security. We'll be supporting "legacy" reads for years to come.
4
1
48
Say it with me, competition is good.
Introducing G4, the world’s most powerful benchtop sequencing platform. Built with novel high-performance chemistry and advanced engineering to deliver accuracy, speed and flexibility, G4 has the ability to power a wide range of
#genomic
applications.
6
42
149
4
2
47
We’re Bioinformaticians that value the art of Software Engineering. Join us if you feel the same.
#Bioinformatics
Join us
@fulcrumgenomics
!
Nowhere else can you have access to so many diverse and challenging
#Bioinformatics
and
#Engineering
problems across so many clients and domains:
‣ 10+ 5 year clients
‣ 100+ projects completed
‣ 50+ clients served
2
28
79
0
13
45
Validate your VCFs before you publish them.
GIAB/NIST and Platinum Genomes VCFs don’t strictly follow the spec.
I’m writing my own tool because I’m a masochist. Early testers are appreciated as I add more features and squash bugs.
#Bioinformatics
2
19
47
There are a lot of positive outcomes in the above process. Folks learn by doing, the software was written to solve an immediate problem, a diversity of ideas and solutions…
But let’s not pretend there isn’t a Bioinformatics SW graveyard, and a lot of it comes down to money.
5
2
47
First release of pybedlite by
@fulcrumgenomics
team member
@natproach
#Python
#Bioinformatics
Motivation was to have a simple way to read and write BED files without having to import the world.
* Read the docs:
* Github:
0
14
46