Amino acid composition predicts prion activity

Abstract

Many prion-forming proteins contain glutamine/asparagine (Q/N) rich domains, and there are conflicting opinions as to the role of primary sequence in their conversion to the prion form: is this phenomenon driven primarily by amino acid composition, or, as a recent computational analysis suggested, dependent on the presence of short sequence elements with high amyloid-forming potential. The argument for the importance of short sequence elements hinged on the relatively-high accuracy obtained using a method that utilizes a collection of length-six sequence elements with known amyloid-forming potential. We weigh in on this question and demonstrate that when those sequence elements are permuted, even higher accuracy is obtained; we also propose a novel multiple-instance machine learning method that uses sequence composition alone, and achieves better accuracy than all existing prion prediction approaches. While we expect there to be elements of primary sequence that affect the process, our experiments suggest that sequence composition alone is sufficient for predicting protein sequences that are likely to form prions. A web-server for the proposed method is available at http://faculty.pieas.edu.pk/fayyaz/prank.html, and the code for reproducing our experiments is available at http://doi.org/10.5281/zenodo.167136.

AMYCO: evaluation of mutational impact on prion-like proteins aggregation propensity

“…pRANK is a novel multiple-instance machine learning method aimed to predict prion propensity based on amino acid composition alone [18]. We compared the performance of AMYCO and pRANK web servers in predicting the impact of mutations on human hnRNPA2 aggregation propensity (Fig.…”

An integrative approach to predicting the functional effects of small indels in non-coding regions of the human genome

“…We used a support vector machine (SVM) [18, 19] as our binary classifier, as SVMs have produced highly accurate classifiers for a variety of bioinformatics domains – see, e.g. [2, 15, 20, 21]. Kernel methods such as SVMs can easily handle structured data, such as strings and graphs, which are abundant in bioinformatic applications.…”

In silico Characterization of Human Prion-Like Proteins: Beyond Neurological Diseases

“…The accumulated knowledge on the determinants of yeast prions conformational conversion has provided strong stimuli for the development of bioinformatics tools to uncover new PrLDs in other organisms (Michelitsch and Weissman, 2000; Harrison and Gerstein, 2003; Toombs et al, 2012; Espinosa Angarica et al, 2014; Lancaster et al, 2014; Afsar Minhas et al, 2017; Batlle et al, 2017c). Previous screenings for PrLDs in the human proteome have targeted the characteristic compositional bias of these protein regions (An and Harrison, 2016).…”