Using the Reactome Graph Database to generate pathway fingerprints for cancer-related genes, and using the fingerprints to gain insights into cancer, comorbidity, and cancer gene networks

Comput Biol Chem. 2025 Sep 19;120(Pt 1):108686. doi: 10.1016/j.compbiolchem.2025.108686. Online ahead of print.

ABSTRACT

The Reactome Graph Database, which describes biological processes in 28 top-level pathways and facilitates efficient data- traversal and retrieval, has been used in a novel way to generate pathway fingerprints for 862 cancer driver- and 324 comorbidity-associated- genes; the latter set of genes cause Mendelian Diseases (MDs) which are comorbid with cancer. A pathway fingerprint presents pathway information about a gene in a novel way. It lists all pathways in which the gene participates, top-level-pathway-wise, the latter in decreasing order of importance. Data related to the above two sets of genes have also been mined from cBioPortal, OMIM, published literature on cancer-MD comorbidity and the NCI-compiled list of targeted therapies approved for cancers. Pathway fingerprints have connected- and added value to- data mined from the different repositories, and yielded insights into cancer and cancer-MD comorbidity. (1) Driver mutations in eight cancer datasets from cBioPortal, and allelic variants (AVs) from OMIM, were extracted for the cancer driver genes; the AVs were searched for among the driver mutations. Only by using the pathway fingerprints of driver genes was it possible to discern that the majority of cancer genes, which recruit AVs as somatic cancer mutations, play their most important role in the Signal Transduction top-level pathway. (2) OMIM-catalogued-, deleterious- AVs in comorbidity-associated genes also occur as somatic mutations in cancers, though infrequently. While in the study their occurrence appears not to be random, in the literature they are regarded as random passenger mutations. In over half the mutations, R is the mutated residue, R→TER mutations being the most frequent. (3) A comparison of the pathway fingerprints of comorbidity-associated- and cancer- genes has shown that the largest number of shared pathways between the two sets of genes belongs to the Signal Transduction top-level pathway. Thus, mutated comorbidity-associated genes help cancer by adding to the burden of aberrant signal transduction. (4) Comorbidity-associated genes which cause Mendelian Inborn Errors of Metabolism tend not to share their disease-causing pathway with cancer genes while contributing to comorbidity. (5) Examining gene targets used in targeted cancer therapies, along with co-occurring genes of each target, yielded novel drug combinations for combination therapy; e.g., Desmopressin as an adjuvant for Dasatinib treatment. (6) Examining the pathway fingerprints of mutated driver genes in each cancer has shown that Signal Transduction and Gene expression are the most-affected top-level pathways in cancers, followed by Cell-Cycle, Developmental Biology and Immune System. (7) Targets for drug design have been identified from the set of genes causing the most prominent effects in the above top-level pathways. (8) Gene networks have been generated which show that: (i) the same source gene, in the same top-level pathway, interacts with similar target genes in different cancers; (ii) if the same source gene, in the same cancer, acts in different top-level pathways, one may identify top-level pathways that are connected by the source gene; (iii) if, in the same top-level pathway, in the same cancer, different source genes act, one may identify source genes that are connected.

PMID:41252800 | DOI:10.1016/j.compbiolchem.2025.108686