|
Activity Number:
|
55
- Statistical methods for data from single cell technologies
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Sunday, August 8, 2021 : 3:30 PM to 5:20 PM
|
|
Sponsor:
|
Section on Statistics in Genomics and Genetics
|
|
Abstract #319065
|
|
|
Title:
|
SpotClean Adjusts for Spot Swapping in Spatial Transcriptomics Data
|
|
Author(s):
|
Zijian Ni* and Aman Prasad and Shuyang Chen and Lisa Arkin and Richard Halberg and Beth Drolet and Michael Newton and Christina Kendziorski
|
|
Companies:
|
UW-Madison and Department of Dermatology, UW-Madison and University of Wisconsin-Madison and Department of Dermatology, UW-Madison and Department of Medicine, UW-Madison and Department of Dermatology, UW-Madison and University of Wisconsin, Madison and UW-Madison
|
|
Keywords:
|
spatial transcriptomics;
decontamination;
quality control;
statistical methods
|
|
Abstract:
|
Spatial transcriptomics has emerged as a revolutionary tool that allows us to address scientific questions that were elusive just a few years ago. In a spatial transcriptomics experiment, genome-wide gene expression is measured from thin tissue sections placed on a slide containing spatially resolved nucleotide barcodes that bind mRNA. The barcodes are placed onto a grid of 'spots', where typically at each spot the mRNA abundances from relatively few cells are measured using oligonucleotide barcodes that encode the spot's spatial location. Unfortunately, there is considerable contamination in a spatial transcriptomics experiment due to mRNA from one spot binding barcodes at other spots, an artifact we refer to as 'spot swapping'. As we will show, spot swapping adversely affects downstream analyses. To address this, we developed SpotClean, a statistical approach to adjust for contamination due to spot swapping. Results from simulation, case study, and experimental data from a chimeric experiment demonstrate advantages in multiple settings.
|
Authors who are presenting talks have a * after their name.
Back to the full JSM 2021 program
|
