Even though there are some 20,000 genes in the human genome, the same ones capture the attention of researchers over and over again, the Atlantic reports. It notes that it's not necessarily because those genes are important or interesting — though some are — but because those are the genes that could be studied with older investigation techniques.
Researchers from Northwestern University examined the qualities of genes that are intensely studied. As they report in PLOS Biology, they developed a machine-learning tool that could predict the number of publications about a gene based on just 15 traits, such as the size of the gene, its expression in certain tissues, and its tolerance of loss-of-function mutations. But as the Atlantic notes, other traits that influence whether a gene is deeply studied — such as its size, whether it is highly expressed, or if it produces a secreted protein — stem from the field's history.
The Northwestern researchers also found that early career investigators who focus on lesser-studied genes are less likely to become independent researchers, illustrating the pressure to study what's been studied already.
In their paper, the researchers call for new funding mechanisms that focus on "nonredundant and likely highly unpredictable research directions."
"If we don't take targeted approaches to incentivize the study of unstudied genes, the system is not going to change," senior author Luis Amaral tells the Atlantic.