Figure 2.

Mouse AML reflects the mutational landscape of human AML. (A–C) The number of nonsynonymous mutations (A), insertion and deletion events (B), and the number of genes harboring nonsynonymous, insertion, or deletion events in leukemic mice and human AML patients (C). Horizontal bars represent mean. (D) A stacked plot of the genes that are mutated in ≥2% of human AML samples and ≥8% of mouse AML samples. 67.90% of genes mutated in leukemic mice are also mutated in human AML patients, whereas only 5.07% of genes with a protein-coding mutation in AML patients are also mutated in leukemic mice. This enrichment is significant (hypergeometric P ≤ 4.26 × 10−20). (E) Of the genes that harbor protein-coding mutations in both leukemic mice and human AML patients, the fraction of protein domains that are mutated in both species (red) or in only one species (blue) is depicted. The color intensity increases with the number of mutations in a given protein domain. 65.07% of these protein domains are mutated in both species, demonstrating enrichment (hypergeometric P ≤ 4.23 × 10−3). Domains similarly affected are recurrently mutated (Spearman correlation of domain p-values r = 0.53, binomial P ≤ 2.73 × 10−8). (F) The frequencies with which various protein classes are mutated in t(8;21)+ mouse and human AML are not significantly different (P = 0.327, a transformed value of the Pillai-Bartlett statistic from a MANOVA model).

or Create an Account

Close Modal
Close Modal