The Epigenetic Dimension of Protein Structure Is an Intrinsic Weakness of the AlphaFold Program

Abstract

One of the most important lessons we have learned from sequencing the human genome is that not all proteins have a 3D structure. In fact, a large part of the human proteome is made up of intrinsically disordered proteins (IDPs) which can adopt multiple structures, and therefore, multiple functions, depending on the ligands with which they interact. Under these conditions, one can wonder about the value of algorithms developed for predicting the structure of proteins, in particular AlphaFold, an AI which claims to have solved the problem of protein structure. In a recent study, we highlighted a particular weakness of AlphaFold for membrane proteins. Based on this observation, we have proposed a paradigm, referred to as “Epigenetic Dimension of Protein Structure” (EDPS), which takes into account all environmental parameters that control the structure of a protein beyond the amino acid sequence (hence “epigenetic”). In this new study, we compare the reliability of the AlphaFold and Robetta algorithms’ predictions for a new set of membrane proteins involved in human pathologies. We found that Robetta was generally more accurate than AlphaFold for ascribing a membrane-compatible topology. Raft lipids (e.g., gangliosides), which control the structural dynamics of membrane protein structure through chaperone effects, were identified as major actors of the EDPS paradigm. We conclude that the epigenetic dimension of a protein structure is an intrinsic weakness of AI-based protein structure prediction, especially AlphaFold, which warrants further development.

Tld1 is a novel regulator of triglyceride lipolysis that demarcates a lipid droplet subpopulation

“…TOPCONS (Tsirigos et al, 2015) and TM AlphaFold (Dobson et al, 2023) also predicted a membrane-embedded topology for the hydrophobic HR2 sequence ( Fig S3M ). Although they were very similar, the RoseTTAFold structure was selected for further simulations as it has been demonstrated to better predict membrane structures (Azzaz et al, 2022; Hegedűs et al, 2022).…”

“…TOPCONS (Tsirigos et al, 2015) and TM AlphaFold (Dobson et al, 2023) also predicted a membrane-embedded topology for the hydrophobic HR2 sequence ( Supplemental Fig 2M ). Although they were very similar, the RoseTTAFold structure was selected for further simulations as it has been demonstrated to better predict membrane structures (Azzaz et al, 2022; Hegedűs et al, 2022). Tld1 HR2 was then embedded into each lipidic environment deep enough to enable the charged residues at the top of the hairpin (Arg61, Asp90, Asp93, Arg100) to be surface oriented ( Fig 4A, Supplemental Fig 2C ).…”

Before and after AlphaFold2: An overview of protein structure prediction

“…In 67% of a dataset tested, AlphaFold prediction resembled holo form and the proteins were less predictable when the conformational differences between apo and holo forms increased (Saldaño et al, 2022). Moreover, Azzaz et al (2022) demonstrated that structure prediction of membrane proteins by AlphaFold is not reliable, mainly because it presents inconsistencies in the location of the transmembrane domains. They stress that the protein environment influences the amino acid sequence, imposing folding constraints.…”

Structural Basis of Botulinum Toxin Type F Binding to Glycosylated Human SV2A: In Silico Studies at the Periphery of a Lipid Raft

“…Building of hSV2Ag models: The full length of the hSV2A receptor was built via ab initio modeling in Robetta ( ; accessed on 1 November 2022), using the RosseTTAFold method. This modeling method was selected for its accuracy in respecting topology of membrane proteins [ 12 , 13 ]. Amino acids 1 to 161, which represent the intracellular domain of hSV2A, were removed due to their intrinsic disorder.…”