Additional file 3: of CellMapper: rapid and accurate inference of gene expression in difficult-to-isolate cell types
Year: 2016
Type: article
Abstract: Algorithm development, part 1: Performance evaluation of five prospective algorithms using TiGER tissue genes as a gold standard [55], compared to the final algorithm CellMapper. Tukey boxplots show the change in area under the precision recall curve (AUPR) for each tissue, relative to the AUPR achieved by the best-performing prospective algorithm for that tissue. While all five prospective algorithms performed poorly relative to the others in several tissues, CellMapper achieved the highest AUPR in 25 out of 30 tissues and was always within 20Â % AUPR of the best method. This analysis was for algorithm development (see Additional file 1): the prospective algorithms were not originally developed to identify cell type-enriched or tissue-enriched genes, but we tested them in this application because they have been effective using 1â 2 query genes in other contexts, such as finding genes in co-regulated biological pathways (e.g. similar GO terms). MEM multi experiment matrix [44], SPELL Serial Patterns of Expression Levels Locator [45], GR Gene Recommender [56], MI mutual information. (PDF 11 kb)
Source: Figshare
Language: English
Subfield: Molecular Biology
Domain: Life Sciences
Open Access status: green
OpenAlex ID: W7137927252