Presenting on "Accurate and interpretable prediction of drug resistance in M. tuberculosis using deep learning" at the Mathematics and Statistics of Genomic Epidemiology workshop. Thanks to the organizers for the opportunity to present. @BIRS_Mathbirs.ca/events/2022/5-day-wo…
Happy to share that our survey of the current interpretable methods for deep learning models in genomics is finally out in Nature Reviews Genetics! 1/7
nature.com/articles/s41576-0…
Maxwell Libbrecht has joined our Editorial Board, and will be handling papers in our Genomics section!
You can find out more about Max's research here:
plos.io/3LLuoJb
2\ bacterial pathogen within a single human host has important implications for public health. In this research, we introduce SplitStrains, a novel method grounded in a rigorous statistical model, estimates mixed strains' proportions and deconvolves them.
1\ Einar Gabbassov, Miguel Moreno-Molina, Iñaki Comas, Maxwell Libbrecht, and Leonid Chindelevitch published their work on "SplitStrains, a tool to identify and separate mixed Mycobacterium tuberculosis infections from WGS data". Since the occurrence of multiple strains of a
Mohammad Sadegh Saberian defended his MSc thesis on Using DNNs in multiple instance learning to identify putative treatments effective against SARS-CoV-2 based on morphological analysis. Congratulations, Mohammad Sadegh!
1\ Kevin Dsouza, and Adam Li submitted a paper on " Latent representation of the human pan-celltype epigenome through a deep recurrent neural network". In this research, we propose Epi-LSTM, a deep long short-term memory (LSTM) recurrent neural network autoencoder to capture the
New preprint from my student Faezeh Bayat on variance-stabilized units for sequencing-based genomics assays! We found that currently-used log and arcsinh transformations do not correctly stabilize variance and developed a method called VSS that does so. biorxiv.org/content/10.1101/…