# DropLasso: A robust variant of Lasso for single cell RNA-seq data

## TL;DR

DropLasso is a new regularization method tailored for single-cell RNA-seq data, effectively handling dropout noise to improve molecular signature learning compared to existing techniques.

## Contribution

It introduces DropLasso, extending dropout regularization to sparse linear models, specifically designed for noisy scRNA-seq data, and clarifies its relation to elastic net.

## Key findings

- DropLasso outperforms standard regularizations on simulated data.
- DropLasso shows promising results on real scRNA-seq datasets.
- The method effectively handles dropout noise in single-cell data.

## Abstract

Single-cell RNA sequencing (scRNA-seq) is a fast growing approach to measure the genome-wide transcriptome of many individual cells in parallel, but results in noisy data with many dropout events. Existing methods to learn molecular signatures from bulk transcriptomic data may therefore not be adapted to scRNA-seq data, in order to automatically classify individual cells into predefined classes. We propose a new method called DropLasso to learn a molecular signature from scRNA-seq data. DropLasso extends the dropout regularisation technique, popular in neural network training, to esti- mate sparse linear models. It is well adapted to data corrupted by dropout noise, such as scRNA-seq data, and we clarify how it relates to elastic net regularisation. We provide promising results on simulated and real scRNA-seq data, suggesting that DropLasso may be better adapted than standard regularisa- tions to infer molecular signatures from scRNA-seq data.