Automated Machine Learning for Genomics. Genomics + machine learning. Open science + open source. Collaboration + community. Check here for updates.
DC / CHI / SF / ES
Joined May 2020
- Tweets15
- Following32
- Followers76
- Likes22
Pinned Tweet
Team work 🦾!
Good to have so many smart friends chipping in for this, too many to tag right now but major shout outs to @mbmakarious and @FarazFaghri for coordinating this write up!
Looking forward to helping build an open community for Genomics+ML.
#OpenScience #ML #genomics
GenoML: Automated Machine Learning for Genomics
deepai.org/publication/genom…
by Mary B. Makarious et al. including @mike_nalls
#MachineLearning #ComputerScience
GenoML retweeted
Ever wanted to get a TLDR-style primer on #machinelearning basics and applications in genetics?
Check out @IRPatNIH's #CARD github linked below for a great talk from teammate @HamptonLLeonard.
She made it v v clear for all audiences.
Also, @geno_ml!
github.com/NIH-CARD/basic-su…
GenoML retweeted
Just in time for Parkinson's awareness month... our manuscript centered around multi-omics prediction models for Parkinson's disease (PD) risk, using a custom automated machine learning (ML) package (@geno_ml), is out now in npj Parkinson's disease: nature.com/articles/s41531-0…
Thanks @mbmakarious !
Proud of the hard work these 3 put into the #opensource tools they presented today.
Here is a link to some slides that cover resources including (drive.google.com/file/d/1f9v…):
- genetics focused autoML (@geno_ml)
- user-friendly end-end #GWAS pipeline
- easy *-omic integration
GenoML retweeted
A little thread on applying #machinelearning to #genetics data at-scale ... lately people have been asking "how do you account for population substructure in your models?"
Luckily, we've included some functionality in our #opensource project @geno_ml to deal with this.
GenoML retweeted
Biomedical conferences are funny ...“the future of genomics is machine learning across all data modalities for disease insight” ... that’s the now ... check out @geno_ml and tons of other similar projects on GitHub.
Admins are not ready I guess. 🤣 – at Bethesda, MD
GenoML retweeted
This was a big undertaking from many people. We've systematically developed, tested, and optimized a multi-modal PD classification model using an open science ML framework. Looked at top features, drug databases, making new gene networks. Best for large scale screening follow-up.
Clinical data + genomics + @geno_ml = predictions + insights + some nice plots.
Great work from @mbmakarious and teammates ... strong first chapter for her thesis!
Proud to work on fun/interesting things like this with gang from @DataTecnica and @SingletonNeuro labs.
First major use case, more to come soon.
Fully transparent / reproducible prediction workflow for Parkinson’s disease.
Give our autoML code a test run, join the community, suggestions and ideas always appreciated!
#OpenScience #MachineLearning #genomics #parkinsonsdisease
Multi-Modality Machine Learning Predicting Parkinsons Disease biorxiv.org/cgi/content/shor… #bioRxiv
New things happening very soon ... just waiting on that update. Maybe a trip to the model zoo or improved handling of imbalanced data or possibly a major pipeline speed up? Ask @mbmakarious later.
GenoML retweeted
Do you have expertise in advanced microscopy and cell culture, and want to join a cutting-edge iPSC initiative in the neurodegeneration space? Open position in the iPSC Neurodegenerative Diseases Initiative (iNDI), the largest-ever genome-engineering project to date! (1/2)
ALT Graphical overview of iNDI
Jeff and Mary!
@JeffJonggeol and @mbmakarious great lecture on democratizing data+tools and increasing diversity in #openscience at the @pdgenetics virtual conference. Challenging systemic problems in science. I'm just lucky they let me help them with things (sometimes) 🤓.
