EVALUATING THE COMPLEXITY OF GENE COEXPRESSION ESTIMATION FOR SINGLE-CELL DATA

Abstract

With the rapid advance of single-cell RNA sequencing (scRNA-seq) technology, understanding biological processes at a more refined single-cell level is becoming possible. Gene coexpression estimation is an essential step in this direction. It can annotate functionalities of unknown genes or construct the basis of gene regulatory network inference. This study thoroughly tests the existing gene coexpression estimation methods on simulation datasets with known ground truth coexpression networks. We generate these novel datasets using two simulation processes, NORmal-To-Anything (NORTA) and Single-cell ExpRession of Genes In silicO (SERGIO), that use the parameters learned from the experimental data. We demonstrate that these simulations better capture the underlying properties of the real-world single-cell datasets than previously tested simulations for the task. Our performance results on tens of simulated and eight experimental datasets show that all methods produce estimations with a high false discovery rate, potentially caused by high sparsity levels in the data. Finally, we find that commonly used preprocessing approaches, such as normalization and imputation, do not improve the coexpression estimation. Overall, our benchmark setup contributes to the coexpression estimator development, and our study provides valuable insights for the community for single-cell data analyses.