WCCI – IJCNN 2008 Special Session
NN10 -
Analysis of Gene and Protein Expression Data
Organizers:
Vladimir
Kuznetsov, Genome Institute of
Paper submission deadline: December 1, 2007
For instruction on paper submission, visit WCCI website: http://www.wcci2008.org/
Synopsis
Advances in high throughput
technologies, such as microarrays, sequences-based DNA-protein complexes, and mass
spectrometry methods, and the availability of human and other complex genome
sequences now allow scientists to identify gene expression profiles, gene copy
numbers, transcription factor binding sites (TFBS), regulatory pathways,
macromolecular interaction networks at genome scale. Furthermore,
computational approaches are combined to make inferences on complex pictures of
basic biological phenomena such as cancer progression, stem cell
differentiation, etc.
To infer such phenomena,
researchers have widely used computing paradigms such as feed-forward neural
networks, self-organizing feature maps, SVM, independent component analysis,
genetic algorithms, etc., to predict essential gene expression patterns, gene
and protein modules, DNA-protein, protein-protein, RNA-protein networks.
This had led to a numerous approaches for data analysis and data mining, and
web servers providing useful classifications and predictions tools for gene and
protein expression analysis. However, novel neural network based algorithms and
their hybrids with computer simulations and statistic-based approaches, which
are capable of handling diverse high throughput expression data for feature
detection, feature selection, pattern recognition, and evolutionary analysis,
effectively are urgently required. The subsequent challenge is quantitative and
integrative analyses, and adequate interpretation of voluminous data having
potentially low signal to noise ratio, high dimension and essential
incompleteness of genome-scale datasets.
We invite papers dealing with all
aspects of computational analysis and modeling of gene expression,
transcription control (genome, transcriptome, and proteome complexity), mass
spectra, prediction and modeling of different types of macromolecular interaction
networks:
Areas of
interests are but not limited to as follows:
– Preprocessing and de-noising of
data
– Techniques for feature extraction
and gene selection
– Clustering and identifying
co-expressed genes
– Identification gene signatures
and co-regulatory gene patterns
– Prediction of binding sites and
regulatory modules at genome scale level
– Predication and analysis
microRNAs, their targets, and antisense interactions
– Identification and prediction
direct gene targets for TF and their combinations
– Analysis of spatio-temporal gene
expression patterns and finding gene regulatory networks.
– Analysis of normal and disease
pathways
– Optimization and automation of
pattern recognition protocols and methods for complex data visualization
This
specials session is organized by IAPR Technical Committee on Pattern
Recognition for Bioinformatics (TC-20).
Technical
Committee
TBD