Identifying optimal feature transforms for classification and prediction in biological systems: recovering receptive field vectors from sparse recordings

File Description SizeFormat 
final.pdfAccepted version3.5 MBAdobe PDFView/Open
Title: Identifying optimal feature transforms for classification and prediction in biological systems: recovering receptive field vectors from sparse recordings
Authors: Nikolic, K
Evans, B
Item Type: Conference Paper
Abstract: With biological systems it is often hard to adequately sample the entire input space. With sensory neural systems this can be a particularly acute problem, with very high dimensional natural inputs and typically sparse spiking outputs. Here we present an information theory based approach to analyse spiking data of an early sensory pathway, demonstrated on retinal ganglion cells (RGC) responding to natural visual scene stimuli (Katz et al., 2016). We used a non-parametric technique based on the concept of mutual information (MI), in particular, Quadratic Mutual Information (QMI). The QMI allowed us to very efficiently search the high dimensional space formed by the visual input for a much smaller dimensional subspace of Receptive Field Vectors (RFV). RFVs give the most information about the response of the cell to natural stimuli. This approach allows us to identify the RFVs far more efficiently using limited data as we can search the complete stimulus space for multiple vectors simultaneously. The RFVs were also used to predict the RGCs’ responses to any natural stimuli. Another suitable area of application of this algorithm is in diagnostic inference. Currently we are adapting the method to be used for identifying the cancer markers in the volatile organic compounds present in exhaled breath. Once the maximally informative features are established they can be used for diagnostic predictions on new breath samples. Preliminary results of the breathomics analysis will be discussed at the conference. There are several other potential applications such as multiclass categorisation for bacterial strains using ISFET arrays for DNA sequencing. This algorithm can be part of a rapid point-of-care device for identifying the specific infectious agents and recommending appropriate antibiotics. Here we will focus on presenting the algorithm using the example of RFVs of RGCs.
Issue Date: 24-Jun-2016
Date of Acceptance: 11-May-2016
Copyright Statement: © 2016 The Authors
Sponsor/Funder: Engineering & Physical Science Research Council (EPSRC)
Commission of the European Communities
Wellcome Trust
Biotechnology and Biological Sciences Research Council (BBSRC)
Biotechnology and Biological Sciences Research Cou
Engineering & Physical Science Research Council (EPSRC)
Funder's Grant Number: EP/H024581/1
Conference Name: International Conference on Machine Learning (ICML), Workshop on Computational Biology
Publication Status: Published
Start Date: 2016-06-24
Conference Place: New York, USA
Appears in Collections:Faculty of Engineering
Electrical and Electronic Engineering

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.

Creative Commonsx