Figure 3 | Scientific Reports

Figure 3

From: Trivial and nontrivial error sources account for misidentification of protein partners in mutual information approaches

Figure 3

Evaluation of optimized MSA concatenations. (A) True positive (TP) rate of random, optimized and native MSA concatenations. (B) Reassessed TP rate of random, optimized and native MSA concatenations by discounting wrong pairings among sequences with Hamming distance within the 20th percentile of the distance distribution. Optimized solutions with TP rate greater than 30% (p = 0.0005) are shown in blue, while optimized solutions with TP rate lower than 30% are shown in red. Random solutions are shown in gray. (C–G) Hamming distance distribution of MSA B, TP rates versus Hamming distance discounts (the 20th percentile is shown with a dashed line), and TP rates of random (rnd) and optimized (opt1–6) solutions for the 20th percentile Hamming distance cutoff shown for representative systems: 3RRL_AB (C), 1EFP_AB (D), 2NU9_AB (E), 3MML_AB (F), and 1TYG_BA (G). This figure was generated using matplotlib v3.1.2 (https://matplotlib.org/ ).

Back to article page