NEW YORK — With the help of deep learning and artificial intelligence-based image analysis, researchers have identified genetic loci related to human skeletal development. The findings help researchers understand why humans are the only bipedal apes and may also elucidate the genetics of many skeletal diseases.
In a study published in Science on Thursday, researchers from the University of Texas at Austin and the New York Genome Center described the new findings.
They started by applying deep learning models to 31,221 full-body X-ray images from 42,284 individuals in the UK Biobank who volunteered to participate in the study. This AI-based approach enabled authors to measure distances between shoulders, knees, ankles, and other points in the participants.
These measurements were compared to each participant's genetic sequence, based on which the authors found 145 loci in the genome linked to the control of skeletal proportions. "These loci are enriched in genes that regulate skeletal development and those associated with rare human skeletal diseases and abnormal mouse skeletal phenotypes," the authors wrote.
The researchers then developed phenotypic and polygenic scores based on which they identified specific associations between hip and knee osteoarthritis and skeletal proportions of the corresponding regions. This helped them understand the genetic risk of many musculoskeletal diseases.
For instance, their findings showed that individuals with a higher hip width-to-height ratio were more likely to develop osteoarthritis and hip pain. Meanwhile, people with higher proportions of femur length to height were more likely to develop arthritis in their knees, knee pain, and other knee problems. They also noted that people with a higher torso length ratio to height were likelier to develop back pain.
"These disorders develop from biomechanical stresses on the joints over a lifetime," said Eucharist Kun, lead author and a researcher at the University of Texas, in a press statement. "Skeletal proportions affect everything from our gait to how we sit, and it makes sense that they are risk factors in these disorders."
The findings have significant implications for understanding human evolution to bipedalism. "While we know from skeletal records that limb proportions of humans increased as we evolved from chimpanzees to humans, what needs to be looked at is what genetic changes allow these things to happen," co-corresponding author Vagheesh Narasimhan, assistant professor at the Department of Integrative Biology at the University of Texas, told GenomeWeb.
The authors wrote that this study provides genomic evidence of evolutionary change in human arm-to-leg and hip-width proportions consistent with notable anatomical changes in these skeletal proportions in the hominin fossil record. "In contrast to cardiovascular, autoimmune, metabolic, and other categories of traits, loci associated with these skeletal proportions are significantly enriched both in human accelerated regions and in regulatory elements of genes differentially expressed in humans and the great apes throughout development," they wrote.