Prediction of amyloid fibril-forming segments based on a support vector machine

Abstract

Background: Amyloid fibrillar aggregates of proteins or polypeptides are known to be associated with many human diseases. Recent studies suggest that short protein regions trigger this aggregation. Thus, identifying these short peptides is critical for understanding diseases and finding potential therapeutic targets.

Aggregation in Protein-Based Biotherapeutics: Computational Studies and Tools to Identify Aggregation-Prone Regions

“…Other methods based on pattern recognition, 3D profiles, and molecular simulations are also emerging. 11,16,21,34,35,41,[58][59][60] To be able to determine which prediction tool performs the best, these tools need to be evaluated on standardized dataset(s) in a statistically rigorous manner. These testing datasets should contain different protein sequences and structures than those used in the training datasets used to develop the tools.…”

“…60 The method predicts amyloidogenic hexapeptides in protein sequences with an overall accuracy of 81%. The authors take into account 41 physicochemical properties for amino acids.…”

Amyloid-based nanosensors and nanodevices

“…The approaches range from special physico-chemical property windows to position-specific sequence matrices, explicit structural models, and combinations thereof. 108,116,[172][173][174][175][176][177][178][179][180][181][182][183][184][185] Many methods base their prediction on the potential of hexapeptides to form amyloid fibres. This characteristic length has been found to cover several of the core motifs for amyloid fiber formation.…”

Predicting the aggregation propensity of prion sequences

“…It is known that short regions of a polypeptide chain are responsible for amyloid fibril formation (Lopez de la Paz and Serrano, 2004;Sanchez de Groot et al, 2005); in this way, Pafig works with a six-residue sliding window and scans for segments that could form fibrils (Tian et al, 2009) (http://www.mobioinfor.cn/pafig/).…”

“…Of the empirical methods, based on experimental or theoretical analysis of amino acid properties and their contribution to amyloid formation, the following algorithms should be highlighted: Chiti and Dobson (Chiti et al, 2003), DuBay et al (DuBay et al, 2004), TANGO (FernandezEscamilla et al, 2004), Tartaglia et al (Tartaglia et al, 2004), Pawar et al (Pawar et al, 2005), AGGRESCAN (Conchillo-Sole et al, 2007), SALSA (Zibaee et al, 2007), Zyggregator (Tartaglia and Vendruscolo, 2008) and Pafig (Tian et al, 2009). The structure-based methods take into account the three-dimensional (3D) structures of welldefined fibrillar conformations of some proteins or peptides.…”