| Method | APC | Re-weighting | Reference | Software package |
---|
Information-based | MI | No | None | [8, 71] | infCalc |
 | VI |  |  | [65] |  |
 | MIj
| Â | Â | [8] | Â |
 | MIHmin
| Â | Â | Â | Â |
 | MIw
| Â | seq %id | [9] | DCA |
Direct | DI | Yes | seq %id, pseudocount | Â | Â |
 | DI256
| Â | Â | [68] | Code S1 in [68] |
 | DI32
| Â | Â | Â | Â |
 | DIplm
| Â | seq %id | [72] | plmDCA |
 | PSICOV |  | Blosum, pseudocount | [14] | PSICOV |
Phylogenetic | CMPcor
| No | Downsampling | [10] | CoMap |
 | CMPchg
| Â | Â | [2] | Â |
 | CMPvol
| Â | Â | Â | Â |
 | CMPpol
| Â | Â | Â | Â |
- Coevolution methods benchmarked fall into three categories. Information-based methods: MI: mutual information [71], VI: variation of information [65], MIj: MI divided by alignment column-pair entropy, MIHmin: MI divided by minimum column entropy [8], MIw: MI with adjusted amino acid probabilities. Direct methods: DI: direct information—MI with re-estimated joint probabilities [9], DI256, DI32: DI using Hopfield-Potts for dimensional reduction (256 and 32 patterns respectively) [68], DIplm: Frobenius norm of coupling matrices in 21-state Potts model using pseudolikelihood maximization [72], PSICOV: sparse inverse covariance estimation [14]. Phylogenetic methods: CoMap P-values for four analyses CMPcor: substitution correlation analysis [10], CMPpol for polarity compensation, CMPchg for charge compensation, CMPvol for volume compensation [2]