xTrimoGene: An Efficient and Scalable Representation Learner for Single-Cell RNA-Seq Data

AO_SCPLOWBSTRACTC_SCPLOWThe advances in high-throughput sequencing technology have led to significant progress in measuring gene expressions in single-cell level. The amount of publicly available single-cell RNA-seq (scRNA-seq) data is already surpassing 50M records for human with each record measuring 20,000 genes. This highlights the need for unsupervised representation learning to fully ingest these data, yet classical transformer architectures are prohibitive to train on such data in terms of both computation and memory. To address this challenge, we propose a novel asymmetric encoder-decoder transformer for scRNA-seq data, called xTrimoGene, which leverages the sparse characteristic of the data to scale up the pre-training. This scalable design of xTrimoGene reduces FLOPs by one to two orders of magnitude compared to classical transformers while maintaining high accuracy, enabling us to train the largest transformer models over the largest scRNA-seq dataset today. Our experiments also show that the performance of xTrimoGene improves as we increase the model sizes, and it also leads to SOTA performance over various downstream tasks, such as cell classification, perturb-seq effect prediction, and drug combination prediction.