Gabriel Hoffman

Our new software dreamlet scales differential expression analysis for single cell data: doi.org/10.1101/2023.0… Checkout the R package: diseaseneurogenomics.github.io/dreamlet

im new to this twitter thing

@David_McGaughey In short, yes. But each of these steps is particularly challenging, either computationally or statistically, for large-scale single cell datasets. dreamlet uses a precision weighted linear mixed model reformulation of voom, weighting based on read count & cell number.

@g_e_hoffman Is it accurate to say the approach is pseudo bulk, normalize, voom?

Our new software dreamlet scales differential expression analysis for single cell data: doi.org/10.1101/2023.0… Checkout the R package: diseaseneurogenomics.github.io/dreamlet

@PeteHaitch @lcolladotor Hi Peter, Circling back to this, the data and software are now public: twitter.com/g_e_hoffman/st…

@lcolladotor @g_e_hoffman Thanks, Leo!

Gabriel Hoffman @g_e_hoffman is presenting #dreamlet 📦 motivated by a very large #snRNAseq dataset (8.8 million cells, over 1,600 donors with technical replicates, over 500 technical batches). Aka a 160 GB hdf5. gz 😱🤯🤩 Expands #variancePartition, #dream, #BioC2022 1/2

@PeteHaitch @lcolladotor If you're looking for large-scale public scRNA-seq datasets: doi.org/10.1038/s41588… nature.com/articles/s4158… ncbi.nlm.nih.gov/geo/query/acc.… biorxiv.org/content/10.110…

@lcolladotor @g_e_hoffman Do you know what the dataset is and is it public?

@PeteHaitch @lcolladotor We are still generating, QCing and doing our first analysis on the dataset: adknowledgeportal.synapse.org/Explore/Projec…

Our paper on Functional interpretation of genetic variants using deep learning predicts impact on chromatin accessibility and histone modification academic.oup.com/nar/advance-ar…

@kylejdevine I found your wallet at E94th and 2nd in NYC. Is this you? I work at Mount Sinai if you want to pick it up on Monday.