Coevolution + Structure Prediction

Disclaimer: The following list is not intended to be a complete literature of coevolution history but only intended to showcase the milestones towards protein structure prediction from coevolutionary information. Typically the "first" manuscript or abstract@conference that demonstrated the idea. If you see anything missing, please alert us!

1. Using co-mutation data for structure prediction

   • 1964C Yanofsky, V Horn, D Thorpe. Protein structure relationships revealed by mutational analysis. Link

2. Double mutations in MSA and mapping of co-evolving residues on 3D structure

   • 1968"Structure, function and evolution in proteins" (see session III). Link

3. Model to account for co-evolution

   • 1970WM Fitch, E Markowitz. An improved method for determining codon variability in a gene and its application to the rate of fixation of mutations in evolution. Link

4. Using co-variation in MSA to predict 3D structure

   • 1991SA Benner, D Gerloff. Patterns of divergence in homologous proteins as indicators of secondary and tertiary structure: a prediction of the structure of the catalytic domain of protein kinases. Link

5. Using mutual-information to detect coevolution

   • 1991 (for RNA)DKY Chiu, T Kolodziejczak. Inferring consensus structure from nucleic acid sequences. Link
   • 1992 (for Codons)R Farber, A Lapedes, K Sirotkin. Determination of eukaryotic protein coding regions using neural networks and information theory. Link
   • 1993 (for Proteins)B Korber, R Farber, D Wolpert, A Lapedes. Covariation of mutations in the V3 loop of human immunodeficiency virus type 1 envelope protein: an information theoretic analysis. Link

6. Using correlation-matrix to detect coevolution

   • 1994Göbel U, Sander C, Schneider R, Valencia A. Correlated mutations and residue contacts in proteins. Link
   • 1994Taylor WR, Hatrick K. Compensating changes in protein multiple sequence alignments. Link

7. Statistical coupling analysis

   • 1999Lockless SW, Ranganathan R. Evolutionarily conserved pathways of energetic connectivity in protein families. Link

8. Recovering protein structures from sparse contact maps

   • 1997 (contact map)Vendruscolo M, Kussell E, Domany E. Recovery of protein structure from contact maps. Link
   • 1997 (sparse contact map)Skolnick J, Kolinski A, Ortiz AR. MONSSTER: a method for folding globular proteins with a small number of distance restraints. Link
   • 1999 (coevolution)Ortiz AR, Kolinski A, Rotkiewicz P, Ilkowski B, Skolnick J. Ab initio folding of proteins using restraints derived from evolutionary information. Link

9. Corrections

   • 2007 (APC)SD Dunn, LM Wahl, GB Gloor. Mutual information without the influence of phylogeny or entropy dramatically improves residue contact prediction. Link

10. Learning the MRF (or Potts models) for protein sequences to predict contacts

    • 1997Partial correlation coefficient (diagonal-normalized inverse-covariance). Link
    • 1999Approximate partition function using MCMC. Link
    • 2005Heuristic approach (adding edges between high MI). Abstract, Paper
    • 2009Iterative message-passing algorithm (DCA). Link
    • 2009Pseudolikelihood with group-sparsity (GREMLIN). Abstract, Paper
    • 2011Inverse covariance (mfDCA). Link
    • 2012Sparse inverse covariance (PSICOV). Link
    • 2013Pseudolikelihood with L2 and APC (plmDCA, GREMLIN). Link, Link
    • 2018Contrastive divergence (ccmgen). Link
    • 2018Boltzmann machines (bmDCA). Link
    • 2020Masked-language-modeling (single-layer attention). Link
    • 2021Autoregressive (arDCA). Link

11. Learning MRFs across protein families

    • 2020Neural Potts model (explicitly learning Potts models). Link
    • 2020BERT (implicitly learning Potts models). Link
    • 2021MSA Transformer (implicit). Link

12. Using contacts derived from SCA contacts to predict protein structure

    • 2008Fold enumeration. Link

13. Using contacts derived from MRF contacts to predict protein structure

    • 2011Fold enumeration. Link, Link
    • 2011Minimization (DCAfold/EVfold). Link, Link
    • 2014Fragments (Fragfold/Rosetta). Link, Link

14. Extract contacts from MRFs using neural networks

    • 2015CNN on contacts (MetaPSICOV). Link
    • 2016CNN on MRF: protein contact prediction from amino acid co-evolution using convolutional networks for graph-valued images. Link
    • 2017ResNet on contacts (Jinbo Xu). Link
    • 2018ResNet on MRF (AlphaFold). Link, Link

15. Others that need to be investigated

    • Burger L, van Nimwegen E. Disentangling direct from indirect co-evolution of residues in protein alignments. PLoS Comput. Biol. 2010;6:1–17. Link
    • Additional reference: pubmed 3237684