Introduction: Musculoskeletal (MSK) injuries are common. Several genetic loci have been implicated. This study aimed to develop a “machine learning” approach to improve identification of “risk” genes. Bioinformatic tools were used to find new information through assessing the feasibility and performance of the BioOntological-Relationship-Graph-database (BORG) to identify tendon, ligament and inflammatory genes linked to Achilles tendinopathy and anterior cruciate ligament ruptures (ACL).
Subjects and methods: All known human genes (~ 20 000) and whole exome sequencing (WES) genes (~1247) from 20 participants identified in our lab were screened through BORG to identify a list of genes and loci. Prioritised genes were explored in an ACL cohort; controls (CON=232), cases (ACL= 237), NON-contact cases (non-contact mechanism of injury) (NON=137).
Results: Screening of all genes through BORG, yielded ~3500 genes from the human genome and from the WES data, ~415 genes, and all these of genes were linked to tendinopathy and ACL. Fifty-four genes were prioritised from the WES gene list. Two variants in heparin sulphate proteoglycan 2: HSPG2 (rs2291826 A/G, rs2291827 G/A) were explored. The rs2291826 G/G (GG VS A/G+AA) genotype was under-represented in CON (CON=2%; ACL=14%; p<0.001; OR:1.13 95% CI: 0.75-1.71) and NON sub-group (CON=2%; NON=9%; p<0.01; OR:1.08; 95% CI: 0.68-1.70) when males were compared. The rs2291827 A/A (AA VS G/A+GG) genotype was over-represented in CON (CON=7%; ACL=0%; p=0.02, OR; 0.45 95% CI: 0.19-0.97) and similarly the A allele was over-represented in CON (CON=22%; ACL=13%; p<0.001; OR:0.45 95% CI:0.25-0.77) when females were compared.
Conclusion: The results show that HSPG2 gene may be implicated in ACL ruptures and validates the proof of concept of the BORG approach in identifying candidate genes for MSK injuries. We further propose that the BORG tool together with WES sequencing data be collectively used in a genomics approach for MSK injuries.