High throughput prediction of inter-protein coevolution
Sustar, Vid (2018-11-14)
High throughput prediction of inter-protein coevolution
(14.11.2018)
Julkaisu on tekijänoikeussäännösten alainen. Teosta voi lukea ja tulostaa henkilökohtaista käyttöä varten. Käyttö kaupallisiin tarkoituksiin on kielletty.
avoin
Julkaisun pysyvä osoite on:
https://urn.fi/URN:NBN:fi-fe2018112949483
https://urn.fi/URN:NBN:fi-fe2018112949483
Tiivistelmä
Inter-protein co-evolution analysis can reveal in/direct functional or physical protein interactions. Inter-protein co-evolutionary analysis compares the correlation of evolutionary changes between residues on aligned orthologous sequences. On the other hand, modern methods used in experimental cell biological research to screen for protein-protein interaction, often based on mass spectrometry, often lead to identification of large amount of possible interacting proteins. If automatized, inter-protein co-evolution analysis can serve as a valuable step in refining the results, typically containing hundreds of hits, for further experiments. Manual retrieval of tens of orthologous sequences, alignment and phylogenetic tree preparations of such amounts of data is insufficient. The aim of this thesis is to create an assembly of scripts that automatize high-throughput inter-protein co-evolution analysis.
Scripts were written in Python language. Scripts are using API client interface to access online databases with sequences of input protein identifiers. Through matched identifiers, over 85 representative orthologous sequences from vertebrate species are retrieved from OrthoDB orthologues database. Scripts align these sequences with PRANK MSA algorithm and create corresponding phylogenetic tree. All protein pairs are structured for multicore computation with CAPS programme on CSC supercomputer. Multiple CAPS outputs are abstracted into comprehensive form for comparison of relative co-adaptive co-evolution between proposed protein pairs.
In this work, I have developed automatization for a protein-interactome screen done by proximity labelling of B cell receptor and plasma membrane associated proteins under activating or non-activating conditions. Applying high-throughput co-evolutionary analysis to this data provides a completely new approach to identify new players in B cell activation, critical for autoimmunity, hypo-immunity or cancer. Results showed unsatisfying performance of CAPS, explanation and alternatives were given.
Scripts were written in Python language. Scripts are using API client interface to access online databases with sequences of input protein identifiers. Through matched identifiers, over 85 representative orthologous sequences from vertebrate species are retrieved from OrthoDB orthologues database. Scripts align these sequences with PRANK MSA algorithm and create corresponding phylogenetic tree. All protein pairs are structured for multicore computation with CAPS programme on CSC supercomputer. Multiple CAPS outputs are abstracted into comprehensive form for comparison of relative co-adaptive co-evolution between proposed protein pairs.
In this work, I have developed automatization for a protein-interactome screen done by proximity labelling of B cell receptor and plasma membrane associated proteins under activating or non-activating conditions. Applying high-throughput co-evolutionary analysis to this data provides a completely new approach to identify new players in B cell activation, critical for autoimmunity, hypo-immunity or cancer. Results showed unsatisfying performance of CAPS, explanation and alternatives were given.