|
Activity Number:
|
87
- Invited ePoster Session: a Statistical Smörgåsbord
|
|
Type:
|
Invited
|
|
Date/Time:
|
Sunday, July 29, 2018 : 8:30 PM to 10:30 PM
|
|
Sponsor:
|
Biopharmaceutical Section
|
|
Abstract #329089
|
|
|
Title:
|
Zero Counts in Single Cell RNA-Seq Data
|
|
Author(s):
|
Hao Wu and Zhijin Wu*
|
|
Companies:
|
Emory University and Brown University
|
|
Keywords:
|
Single cell;
RNA-seq;
zero-inflation
|
|
Abstract:
|
Single cell RNA sequencing data contain an excessive amount of zero counts. Zero-inflation models have been widely used to account for the substantial point mass at zero of gene expression distribution. It is still being debated whether the source of zero is primarily due to technical dropouts (failure in detection) or true absence of expression. This has led to different strategies of the treatment of zeros -- as evidence for absence or as missing data. We analyze a number of scRNAseq data and found that both sample (cell)-level factors and gene specific factors affect the probability of zero counts. We present a modified generalized model that accounts for the technical variability of detection in individual cells, and present a method for the comparison of the expression probability across populations of cells with adjustment of the technical biases.
|
Authors who are presenting talks have a * after their name.
