TooT Suite: Prediction and classification of membrane transport proteins

A Genome Canada 2017 Bioinformatics and Computational Biology Competition Project

The Toot Suite project is developing tools for predicting and classifying transport proteins in host organisms and microbiomes.

These tools will provide scientists with insight into host-microbiome interactions. For agricultural scientists, the host-microbiome interactions are important for plant and animal health and growth, which are key factors in food productivity.

An Open Science Project following F.A.I.R. Guidelines

A key aspect of the TooT Suite project is to support open, reproducible science. Therefore, we adopt the F.A.I.R. Guidelines for Findable, Accessible, Interoperable, and Reusable project artefacts. This includes our experimental platform based on Boutiques for computation on the clusters of Compute Canada and the Gina Cody School of Engineering and Computer Science. and a git repository holding our datasets, source code, and processing pipelines.

Deliverables

Due Deliverable Description Link
Q4 D.1.1 TooT-SC: Given a fasta file of protein sequences, predict the substrate class that is transported.
TooT-BERT-SC: Classifies transmembrane transporter proteins by substrate using a majority-vote approach with various standard predictors.		 Link
Link
Q4 D.1.4 Computational infrastructure for in silico experiments. Link
Q4 D.4.1 Experimental infrastructure. Link
Q6 D.1.2 TooT-Proteome: Given a fasta file of a proteome, predict the transporters and their properties.
TooT-BERT-T: Predicts transmembrane transporter proteins from protein sequences using ProtBERT-BFD and logistic regression.
TooT-Proteome: Predicts transmembrane transporter proteins and their substrate-specific properties using BERT-based predictors for transporter and substrate classification.		 Link
Link

Link
Q6 D.2.1 and D.3.1 (HPC) TooT-TC: Given a fasta file of protein sequences, predict their TC family and subfamily.
Q8 D.2.2 TooT-All: Given a fasta file of protein sequences, classify against all transporter schemes.
Q8 D.4.2 Experimental infrastructure - version 2. Link (boutiques)
Link (CBRAIN)
Link (biolab VM)
Q12 D.1.3 TooT-BERT-ICAT: Utilizes ProtBERT-BFD to predict and classify inorganic ion-specific transmembrane transporter proteins with high precision and accuracy.
TooT-SS: Given a fasta file of protein sequences, predict the specific substrate that is transported.		 Link
Q12 D.2.3 TooT-All (version 2): Given a fasta file of protein sequences, classify against all transporter schemes.
Q12 D.3.2 TooT-Meta (= TooT-Proteome HPC): Given a fasta file of a meta-proteome, predict the transporters and their properties.