Fig 1: Mutation- and EGF-dependent changes in SHP2-TurboID proximity labeling.(A) Stringent pairwise overlap in hits for all SHP2 variants relative to the TurboID control in unstimulated samples. Hits are defined as proteins with >2-fold enrichment and a p value <0.05 (heteroscedastic, unpaired t-test), and overlap indicates the number of proteins that meet these criteria for both SHP2 variants in the comparison. (B) Same as in (A) but with EGF stimulation. (C) Relaxed pairwise overlap in hits for all SHP2 variants relative to TurboID control in unstimulated samples. For SHP2 variants on the y-axis, a hit was defined as proteins with both >2-fold enrichment and a p value <0.05. For SHP2 variants on the x-axis, a hit was defined as proteins with either >2-fold enrichment or a p value <0.05. (D) Same as (C), but with EGF stimulation. Note that in panels (A–D), the smaller numbers for SHP2Y279C reflect the use of only two replicates for this mutant, due to an instrument error and sample loss. (E) Enrichment of MPZL1 in our dataset for all SHP2 variants relative to the TurboID-only control (+EGF), highlighting significant proximity labeling with some mutants (n = 3 biological replicates for all samples except SHP2Y279C, which had two biological replicates). Yellow indicates a significant difference from the TurboID-only control (fold-change >2, p value <0.05 from a heteroscedastic, unpaired t-test): p = 0.005656 (T42A), p = 0.00580 (E139D), p = 0.00003 (Q510E), and p = 0.00049 (Q510K). (F) Enrichment of proteins with predicted C-SH2 binding sites (top ~30th percentile) by SHP2WT, SHP2T42A, and SHP2R138Q, showing loss of signal for C-SH2-dead SHP2R138Q (n = 3 biological replicates). In panels (E, F), bar heights indicate the mean, and error bars indicate standard deviation. Source data are available online for this figure.