Skip to main content

Table 2 Computational evaluation of peptides selected for experimental verification

From: Searching for universal model of amyloid signaling motifs using probabilistic context-free grammars

Peptide Sequence
ORT49035.1_103_123 VDLRDAKGVQVGDGNVQINRF
AAS80314.1_349_385 SFNNLGSGDQFNTPGGTQNINKGGNEVSGGNFYGSVQF
AEB69175.1_5_29 KSRFDQRGQKVIGQQINVAGDATLP
RDW70414.1_382_421 GAPANNTSNSIQHNNSGSGHQNSGSGQQNIGTNTGSGQQ
Peptide PCFG pHMM PLAAC PASTA2 AmyloG ArchC
FPR\(\downarrow\) FPR\(\downarrow\) FPR\(\downarrow\) [PEU]\(\downarrow\) [0-1]\(\uparrow\) [0-1]\(\uparrow\)
ORT49035.1_103_123 \(\le\) 0.0004 < 0.0004 0.022 − 2.5 0.63 0.379
AAS80314.1_349_385 < 0.0004 < 0.0004 0.0004 − 1.6 0.78 0.555
AEB69175.1_5_29 0.007\(- 0.07\) 0.0004 0.022 − 3.7 0.43 0.488
RDW70414.1_382_421 0.0004 \(0.01{-}0.35\) <0.0004 − 0.9 0.26 0.599
  1. Top: List of peptides selected for experimental verification. Bottom: results of computational methods. For PCFG, pHMM (both with score averaging) and PLAAC we provide FPR against NLReff at which at least a part of the peptide is detected; for PASTA2—Pasta Energy Units; for AmyloGram (AmyloG) and ArchCandy (ArchC)—score ranged from 0 to 1. Arrow indicates if higher or lower value is more positive. Values exceeding the default or suggested thresholds are shown in bold (the thresholds are \(1e{-}2\) for PCFG, pHMM and PLAAC, -5 PEU for PASTA2, 0.5 for AmyloGram, and 0.56 for ArchCandy). For PCFG, we provide the range of values for grammars of different maximum number of symbols; For pHMM—the range of values for profiles made from different accurate MSAs