Page 9 - Cancer Systems Biology: Methods and Protocols

Combinatorial Mutational Patterns in Human Cancers 5

mutations which will not change an amino acid (marked as "coding
silent" in COSMIC). Those mutations typically have little, if any,
impact on the biological function of corresponding proteins and are
uninformative for signaling pathway inference [I1, 15]. Other types
of mutations, such as missense and nonsense point mutations, small
insertions and deletions, frame shifts, gene fusions, and transloca-
tions, ctc., could be counted as effective mutations when performing
exclusivity analysis. Furthermore, the mutations should be detected
based on genome-wide or exome-wide screening efforts ensuring
that all protein-coding genes were covered, to minimize the statisti-
I: the cal bias induced by incomplete sample coverage.
0us!y The step-by-step procedure for data acquisition and criteria of
nario: data quality control, as well as the specific formulae used to calcu-
late the likelihood ratio (LR) and significance level (p-value), are
elucidated in the following sections. Figure 2 illustrates the overall
2ntly procedures of this pipeline for mutational pattern determination.
et of
:erns
curs Mutation
ngle
1ar1o
l in a

ame Quality control
take
enes
man
have Preprocessing
don
has
with

this
cant Determine
med mutational pattern
.s an For Inferring
tion signaling
network
s of ±#k
.. ··J
atic �ft� tr - ·
·
�
ome : {
1on, d •
s to «
onal
the ..........
'."
" .+++++,
able Fig. 2 Schematic of the overall pipeline proposed in this protocol. The specific
ncer steps of text processing, computation, and visualization are provided in
oint Subheading 3