A literature-based approach for curating gene signatures in multifaceted diseases

Abstract

Background and aims The task of identifying a representative and yet manageable target gene list for assessing the pathogenesis of complicated and multifaceted diseases is challenging. Using Inflammatory Bowel Disease (IBD) as an example, we conceived a bioinformatic approach to identify novel genes associated with the various disease subtypes, in combination with known clinical control genes. Methods From the available literature, we used Acumenta Literature Lab TM (LitLab), network analyses, and LitLab Gene Retriever to assemble a gene pool that has a high likelihood of representing immunity-related subtype-specific signatures of IBD. Results We generated six relevant gene lists and 21 intersections that contain genes with unique literature associations to Crohn’s Disease (n = 60), Ulcerative Colitis (n = 17), and unclassified (n = 45) subtypes of IBD. From this gene pool, we then filtered and constructed, using network analysis, a final list of 142 genes that are the most representative of the disease and its subtypes. Conclusions In this paper, we present the bioinformatic construction of a gene panel that putatively contains subtype signatures of IBD, a multifactorial disease. These gene signatures will be tested as biomarkers to classify patients with IBD, which has been a clinically challenging task. Such approach to diagnose and monitor complicated disease pathogenesis is a stepping-stone towards personalized care.

Harnessing the Power of Precision Medicine and Novel Biomarkers to Treat Crohn’s Disease

“…The Personalizing Anti-TNF Therapy in the Crohn’s Disease (PANTS) study, which included 1240 biologic-naive CD patients from the UK, has revealed an intriguing finding that is highly likely to have clinical implications. It discovered that variations in the HLA-DQA1*05 allele are linked to a higher likelihood of developing antibodies against anti-TNF agents and that this immunogenicity can be significantly decreased by using a concurrent immunomodulator [ 67 ]. A future randomized controlled biomarker trial is necessary to determine whether pretreatment testing for HLA-DQA1*05 will improve patient outcomes by aiding physicians in selecting anti-TNF (with or without combination therapies), and whether it is feasible to precheck every patient.…”

Inflammatory Bowel Disease Treatments and Predictive Biomarkers of Therapeutic Response

“…Genetic markers: Similarly, the genetic profiling of markers has shown a positive correlation with predictive response to biological treatment in IBD patients. Most genetic predictive markers are related to cytokines or their receptors and immunoglobulin receptors, including TNF/TNF-receptor genes, ATG16L1 gene, apoptosis genes, NOD2/CARD15 genes, CRP, IL23R and IL12 genes and Fc receptors related genes [175][176][177][178]. For example, genetic variations in TNF-β and TNFRSF1B genes (rs1061624_A-rs3397_T) together with a minor allele (A) polymorphism of TNF gene (rs1800629) could predict a non-responsiveness to anti-TNF (infliximab) therapy in CD patients [179][180][181], while a heterozygous genotype of IL12B-10993 G > C (rs3212217) is positively correlated with non-responsiveness to anti-TNF therapy in UC patients [182].…”

Exploring Molecular Signatures in Spondyloarthritis: A Step Towards Early Diagnosis

“…This is implemented by determining the number of abstracts that mention the gene and the terms within the term list and by ranking the terms based on their association with the genes (called as, product of frequency or PF). It then compares the LPF value of an experimental gene set to values of 1000 random gene sets so as to determine the strength of association, resulting in a score [19,20]. It was used to retrieve literature associated genes for the terms: Psoriatic arthritis (PsA), Reactive arthritis (ReA), Rheumatoid arthritis (RA) and Juvenile arthritis (JA).…”