The Budapest Amyloid Predictor and Its Applications

Abstract

The amyloid state of proteins is widely studied with relevance to neurology, biochemistry, and biotechnology. In contrast with nearly amorphous aggregation, the amyloid state has a well-defined structure, consisting of parallel and antiparallel β-sheets in a periodically repeated formation. The understanding of the amyloid state is growing with the development of novel molecular imaging tools, like cryogenic electron microscopy. Sequence-based amyloid predictors were developed, mainly using artificial neural networks (ANNs) as the underlying computational technique. From a good neural-network-based predictor, it is a very difficult task to identify the attributes of the input amino acid sequence, which imply the decision of the network. Here, we present a linear Support Vector Machine (SVM)-based predictor for hexapeptides with correctness higher than 84%, i.e., it is at least as good as the best published ANN-based tools. Unlike artificial neural networks, the decisions of the linear SVMs are much easier to analyze and, from a good predictor, we can infer rich biochemical knowledge. In the Budapest Amyloid Predictor webserver the user needs to input a hexapeptide, and the server outputs a prediction for the input plus the 6 × 19 = 114 distance-1 neighbors of the input hexapeptide.

The polyglutamine amyloid nucleus in living cells is a monomer with competing dimensions of order

“…As a starting point to uncover possible physical explanations for the dependence on even/odd spacing, we analyzed representative sequences -- 60:Q 3 N, 60:Q 4 N, 60:Q 5 N, 60:Q, 60:N -- with state-of-the-art amyloid predictors (Charoenkwan et al, 2021; Keresztes et al, 2021; Prabakaran et al, 2021). Using their respective default parameters, we found that most predictors failed entirely to detect amyloid propensity among these sequences.…”

Succinct Amyloid and Non-Amyloid Patterns in Hexapeptides

“…We have introduced the Budapest Amyloid Predictor webserver in the work [14] by applying linear Support Vector Machines as the underlying prediction tool [15], and the Waltz dataset [16,17] for training and testing purposes. The Waltz dataset consists of 1415 hexapeptides, from which 514 peptides are experimentally labeled as "amyloidogenic" and 901 hexapeptides as "nonamyloidogenic".…”

“…Therefore, a hexapeptide was represented by a length 6 × 553 = 3318 vector z. We note that this highly redundant representation has given somewhat better predictions than more concise ones [14] and has not caused any difficulties in what follows.…”

“…Recently we built a Support Vector Machine for amyloidogenecity prediction of hexapeptides [14]. The training set was the Waltz dataset [16,17] of experimentally identified 514 amyloidogenic and 901 non-amyloidogenic hexapeptides.…”

“…. , y m were assigned to each hexapeptide, and using this multi-dimensional representation, we have prepared an SVM, called Budapest Amyloid Predictor (abbreviated as BAP), described in detail in [14], and freely available at the address https://pitgroup.org/bap.…”