The increased digitalization of the biological data along with the development of increasingly advanced artificial intelligence algorithms are radically changing the way of approaching the biology of living beings.
Indeed, Deep Learning (DL) methods are generic approaches to learning functional correlations from data without having to specify them beforehand. Their attraction in computational biology is the capacity to generate predictive models without making significant assumptions about underlying processes, which are typically unclear or inadequately characterized. DL is now transforming peptide drug research as a result of both computational breakthroughs and the significant recent increase of digital biological data.
In a recent paper published in Bone Research, Cai and colleagues developed a DL-based oligopeptide discovery technique used an unsupervised DL of the intrinsically disordered regions (IDRs) present in proteins. In particular, they focused on osteogenesis by analysing the IDRs of 171 osteogenesis-annotated proteins. From the DL analysis, they selected 28 potential oligopeptides with predicted pro-osteogenic effect. Among them, the Artificial Intelligence-developed Bone-forming Pentapeptide (AIB5P) was confirmed to promote osteogenesis both in vitro and in vivo.
Overall, the DL-based oligopeptide discovery describe by Cai and colleagues can potentially speed up the development of oligopeptide medicines for the treatment of a wide range of clinical conditions.