Results of both experimental techniques are verified in a complementary manner using the mass conservation calculation.
Analysis of amino acid conservation in GH family sequence alignments were performed with AL2CO program [41], using as parameters a smooth length of 3 residues, BLOSUM62 as scoring matrix, the independent count for sequence weighting scheme and entropy as conservation calculation method [41].
The same sequence conservation calculation procedure was used for another two genome sequence categories: (i) randomly selected intergenic regions and (ii) annotated lncRNAs.
To allow for ambiguity in annotation, if the second or third most common intergenic distance is within 1 of the mode, these gene-pairs are included in the conservation calculation.
To test the functional relevance of our SLiM conservation calculation, we analyzed reported SH2 binding sites in 11 highly-studied RTKs (with greater than 30 interaction partners, according to Hprd), including EGFR, IR, KIT, PDGFRB, IGF-IR, VEGFR2, ERBB2, FGFR1, HGFR, RET and TKR-A.
Using our SLiM conservation calculation, we found that reported SH2 binding sites have significantly higher (p < 0.0001, Mann-Whitney test) ln(CR) scores (which measures the relative conservation of a motif) than those sites that do not bind to SH2 domains, indicating the relative conservation score is an effective distinguishing factor of functional binding SLiMs.
Using our SLiM conservation calculation, we found that SLiMs reported to bind to SH2 domains have significantly higher ln(CR) scores than those SLiMs that do not bind to SH2 domains in both disordered and ordered sequences (Mann-Whitney test p < 0.001 in both cases, Figure 2A).
Within the GeneSV system, we have implemented 2 measures: (sCON) and (eCON) that explore well-known strategies for sequence conservation calculations.
In Table 1, we provide secondary structure element assignments (SSE), solvent accessibility (ACC) scores, predictions if a given position is a part of the epitope (EPI), and results from the conservation calculations (eCON), and (sCON) for FLIC and RefSeq.
Therefore, it is uncertain that the current tools, which are generally based on conservation calculations, can be trusted to screen whole-genome sequencing data in search of driver mutations in PKs.
This study is novel because of the way it combined measurements with mass-conservation calculations.
