paperKB
coga / coga-kb
Help
Sign in

Chunk #37 — Results — Using L1000 data to assess allele function

Source
A Next Generation Connectivity Map: L1000 Platform and the First 1,000,000 Profiles.
Embedded
yes

Text

The preceding analyses focused primarily on using CMap to annotate chemical compounds. We next asked whether a similar strategy could be used to annotate the function of an allelic series of genes. Building on our prior results (Berger et al., 2016), we sought to determine whether the CMap could distinguish the downstream consequences of overexpression of cDNAs harboring particular somatic mutations observed in human tumors. For example, the ubiquitin ligase FBXW7 is a negative regulator of MYC protein expression. As expected, CMap showed that overexpression of wild-type FBXW7 strongly connected to knock-down of MYC. In addition, overexpression of 6 cancer-associated alleles (I347M, V464E, R465C, R465H, A502V, and R505C) all lost this connection to MYC loss-of-function, whereas 4 other alleles retained connectivity to MYC knockdown (Figure 7A, lower panel). Examination of the substrate-bound FBXW7 crystal structure (Hao et al., 2007) indicated that the mutations predicted by the CMap to be damaging map to the substrate-recognition pocket, whereas the non-damaging alleles do not (Figure 7A, upper panel).