eweitz / ideogram

Chromosome visualization for the web
https://eweitz.github.io/ideogram
Other
288 stars 72 forks source link

Use gene cache to detect if interaction node is gene #274

Closed eweitz closed 3 years ago

eweitz commented 3 years ago

This uses the new gene cache (#272) to detect if an interaction node represents a gene. It fixes a bug in the related genes kit, e.g. when searching e.g. PIK3R1 as shown below. The bug has INPP5F reported as interacting with PIK3R1 -- but the two genes don't actually interact, thus the tooltip unexpectedly shows no pathways.

Bug:

Bug_in_PIK3R1_related_genes_ideogram_2021-06-04

Fix:

PIK3R1_related_genes_ideogram_2021-06-04

The problem occurred when WikiPathways returned nodes for non-genes, and Ideogram would query MyGene.info using the non-gene labels as gene symbols, e.g. https://mygene.info/v3/query?q=symbol:PRKCA%20OR%20symbol:GRB2%20OR%20symbol:PIK3CA%20OR%20symbol:AKT1%20OR%20symbol:VAV1%20OR%20symbol:PIP3%20OR%20symbol:PTPN11%20OR%20symbol:PIK3CD%20OR%20symbol:CRK%20OR%20symbol:SRC%20OR%20symbol:CD28%20OR%20symbol:CBL%20OR%20symbol:SHB%20OR%20symbol:IL5RA%20OR%20symbol:GNB1%20OR%20symbol:hsa-mir-15b%20OR%20symbol:IL1B%20OR%20symbol:LAT%20OR%20symbol:GAB1%20OR%20symbol:IRS2%20OR%20symbol:SH2B1%20OR%20symbol:LCK%20OR%20symbol:PI(3,4,5)P3%20OR%20symbol:AKT3%20OR%20symbol:PIK3CB%20OR%20symbol:AP2A1%20OR%20symbol:PIP2%20OR%20symbol:TRB%20OR%20symbol:AKT2%20OR%20symbol:PDK1%20OR%20symbol:Invasion%20OR%20symbol:Osteoblast%20OR%20symbol:PIK3C2B%20OR%20symbol:PIK3%20OR%20symbol:PRKCZ&species=9606&fields=symbol,genomic_pos,name&size=20.

In that case, some non-gene "symbols" would convert to gene symbols in MyGene.info, and be spuriously displayed as interacting genes in the related genes ideogram.

By using the complete list of known genes from the gene cache, Ideogram can not robustly detect if a WikiPathways interaction node represents a gene.

coveralls commented 3 years ago

Coverage Status

Coverage decreased (-0.1%) to 89.155% when pulling ff190f2f794ef6bf10e2316b818d14e5435fb5a7 on gene-check-interaction into f59169643ae562acd102c78e96883a6521641410 on master.