nealgravindra Goto Github PK

cluster2celltype

Using biological priors (previously selected important features or "markers"), assign cell types to clusters based on differential discriminability of cell type markers

condense

CONditional DENSity Embedding for unsupervised characterization of gene dynamics

covid_severity_index

CSI developed in the early days, published in https://www.annemergmed.com/article/S0196-0644(20)30588-6/fulltext

cpa

The Compositional Perturbation Autoencoder (CPA) is a deep generative framework to learn effects of perturbations at the single-cell level. CPA performs OOD predictions of unseen combinations of drugs, learns interpretable embeddings, estimates dose-response curves, and provides uncertainty estimates.

diffuse_profile

How to diffuse a profile into a shared space for recommendations? Should we use supervised learning and feature attribution or variance metrics for perturbation profiles? How to evaluate this approach?

distributional_encoder

use of a permutation invariant network to encode probability distributions

dlmodules

3 month-ish thematic modules to keep up with new developments in other fields -- blogging, writing a textbook, implementing code from scratch

ed_mortality

fromscratch

implementations of papers from scratch

gnn_scgexa

GNN-based single-cell gene expression analysis

graphtransormer

harmonized-toolkit

The Harmonized Pre-processing Toolkit for Dementia Bank

lungmap

nalab

new2dev

Playground for new computational implementations in order to appropriately tweak and transition to personal prj dev

pytorch_geometric

Geometric Deep Learning Extension Library for PyTorch

readinglist

need to read more papers gahhh

rl4xai

sars-cov-2_hbec

scgat

Code to reproduce results in Ravindra NG* & Sehanobish A* et al. 2020, "Disease State Prediction From Single-Cell Data Using Graph Attention Networks" to appear in ACM Proceedings. <https://arxiv.org/abs/2002.07128>

scgat_scnd

tutorial for Cell paper

scnd

neurodegeneration and single-cell omics

scsarscov2

self-supervsed_edge_feats

code for "Self-supervised edge features for improved Graph Neural Network training", <arxivlink>

seq2gex

Learning how life's languages are expressed

simpletopics

composite scores for indexed descriptions to target document(s)

single-cell_neurodegeneration

Code to reproduce results of Tejwani L* and Ravindra NG* et al. 2020

stuck_in_plat

for the ave. apex legends player that needs to scheme to rank up

talks

torch_accelerated_skeleton

based on a talk by Sebastian Rascka at CVPR, finally gonna use fabric