Whole-proteome interaction mining

Abstract

The interaction mining approach was demonstrated by building a learning system based on 1,039 experimentally validated protein-protein interactions in the human gastric bacterium Helicobacter pylori. An estimate of the generalization performance of the classifier was derived from 10-fold cross-validation, which indicated expected upper bounds on precision of 80% and sensitivity of 69% when applied to related organisms. One such organism is the enteric pathogen Campylobacter jejuni, in which comprehensive machine learning prediction of all possible pairwise protein-protein interactions was performed. The resulting network of interactions shares an average protein connectivity characteristic in common with previous investigations reported in the literature, offering strong evidence supporting the biological feasibility of the hypothesized map. For inferences about complete proteomes in which the number of pairwise non-interactions is expected to be much larger than the number of actual interactions, we anticipate that the sensitivity will remain the same but precision may decrease. We present specific biological examples of two subnetworks of protein-protein interactions in C. jejuni resulting from the application of this approach, including elements of a two-component signal transduction systems for thermoregulation, and a ferritin uptake network.

Prediction of protein-protein interactions using chaos game representation and wavelet transform via the random forest algorithm

“…The results of the 10-fold cross-validation for four different methods (Bock and Gough, 2003;Martin et al, 2005;Nanni, 2005a,b) for the H. pylori dataset are shown in Table 6. In Bock and Gough's approach Gough, 2001, 2003), several structural and physiochemical descriptors, with SVM as the classifier, were used to predict PPIs.…”

“…In another paper, Nanni (2005b) proposed a PPIs predictor based on linear discriminant classifiers. Results obtained by the 10 cross-validations for the predictor by Bock and Gough (2003) for the H. pylori dataset. See the ''CGR and two related discrete series'' section for further explanation about 10 crossvalidations.…”

High performance set of PseAAC and sequence based descriptors for protein classification

“…(Garg and Gupta, 2008): this dataset consists of 83 SWISS-PROT sequences (40 virulent and 43 nonvirulent protein sequences) such that there are no two sequences that are more than 40% similar. (Garg and Gupta, 2008) (Bock and Gough, 2003): this dataset contains a total of 1882 human protein pairs. Each pair of proteins is labeled as either an interacting pair or a non-interacting pair.…”

“…Each pair of proteins is labeled as either an interacting pair or a non-interacting pair. (Bock and Gough, 2003): this dataset contains a total of 2916 helicobacter protein pairs. Each pair of proteins is labeled as either an interacting pair or a non-interacting pair.…”

A Hybrid Rule-Induction/Likelihood-Ratio Based Approach for Predicting Protein-Protein Interactions

“…Many methods infer interactions from a single type of genomic data. For example, [3] and [4] address the question whether protein interactions can be predicted directly from the primary structure and associated data. Given a database of interacting proteins, they develop a machine learning system (Support Vector Machine) trained to recognise the potential interactions based solely on the primary structure and the associated physicochemical properties.…”