Please use this identifier to cite or link to this item:
Title: Classification between odorant and gustatory receptors: Supervised learning applied to arthropod proteins
Author: Enríquez Romero, Félix Francisco
Tutor: Orengo Ferriz, Dorcas
Others: Calvet Liñan, Laura  
Keywords: odorant receptor
gustatory receptor
supervised learning
Issue Date: 1-Jan-2022
Publisher: Universitat Oberta de Catalunya (UOC)
Abstract: Functional characterisation of proteins is limited and difficult to achieve using automatic systems based on sequence alignments which have not found evidence of the presence of odorant receptors in other arthropods than insects. In this project, a supervised learning based system is developed in order to first classify insect protein sequences functionality as odorant or gustatory receptors and second, identify in non-insect arthropod sequences (annotated as gustatory receptors by alignment based systems) potential sequences that suffered a similar functional divergence as it happened in insects around 440 million years ago. Dataset sequences were obtained from UniProtKB, three different encoding methods were used to train three artificial neural networks. The results obtained confirm the efficiency of the developed model with AUC, precision and F1 score of 0.911, 0.971 and 0.926, respectively. A list of candidate sequences to have a functionality similar to that of odorant receptors in non-insect arthropods was obtained. Downstream analysis regarding functionality of these sequences should be done to corroborate the model predictions. Findings regarding the presence of proteins with similar functionality to odorant receptors in other arthropods would indicate an extraordinary evolutionary convergence.
Language: English
Appears in Collections:Bachelor thesis, research projects, etc.

Files in This Item:
File Description SizeFormat 
fferTFM0122presentationvideo.mp4Presentation Video68.68 MBMP4View/Open
fferTFM0122memory.pdfTFM Memory700.44 kBAdobe PDFView/Open

This item is licensed under a Creative Commons License Creative Commons