Homepage

Description

The data was collected from several sources, including GenProtEC ([Web Link]) and SWISSPROT ([Web Link]). Structure prediction was made by PROF ([Web Link]). Homology search was provided by PSI-BLAST ([Web Link]). The data is in Datalog format. Missing values are not explicit, but some genes have more relationships than others. E. coli genes (ORFs) are related to each other by the predicate ecoli_to_ecoli(EcoliNumber,E-value,Psi-blast_iteration). They are related to other (SWISSPROT) proteins by the predicate e_val(AccNo,E-value). All the data for a single gene (ORF) is enclosed between delimiters of the form: begin(model(EcoliNumber)). end(model(EcoliNumber)). The gene functional classes are in a hierarchy. See [Web Link] (note: the classes may have changed since original data collection). There are two datalog files: ecoli_data.pl and ecoli_functions.pl 1. ecoli_functions.pl Lists classes and ORF functions. Lines are of the following form: class(5,1,1,'Colicin-related functions'). class(5,1,'Laterally acquirred elements'). class(5,'Extrachromosomal'). Arguments are up to 3 numbers (describing class at up to 3 different levels), followed by a string class description. For example: function(ecoli210,7,0,0,'b0217','putative aminopeptidase'). Arguments are ORF number, exactly 3 class numbers, gene name (or blattner number if no gene name), ORF description. 2. ecoli_data.pl Data for each ORF (gene) is delimited by begin(model(ecoliX)). end(model(ecoliX)). where X is the ORF number. Other predicates are as follows (examples): ecoli_orf(ecoliX). % X is ORF number ecoli_mol_wt(176624.1). % float ecoli_theo_pI(5.81). %float ecoli_atomic_comp(c,7940). % {c,h,n,o,s} , int ecoli_aliphatic_index(69.57). % float ecoli_hydro(-0.549). % float sec_struc(1,c,2). % int (start), {a,b,c}, int (length) sec_struc_coil(1,2). % int (start), int (length) sec_struc_beta(1,5). % int (start), int (length) sec_struc_alpha(1,7). % int (start), int (length) sequence_length(255). % int amino_acid_ratio(a,8.9). % amino_acid_char, float amino_acids(ecoli3013,a,70). % ORF_num, amino_acid_char, int amino_acid_pair_ratio(a,a,9.0). % amino_acid_char, amino_acid_char, float amino_acid_pairs(a,a,7). % amino_acid_char, amino_acid_char, int ecoli_to_ecoli(1170,1.0e-105,5). % ORF_num, double (e-value), int (iteration) e_val(o42893,2.0e-99). % accession_number, double (e-value) psi_iter(o42893,5). % accession_number, int (iteration) species(p52494,'candida_albicans__yeast_'). % accession_number, string mol_wt(p52494,104022). % accession_number, int classification(p52494,candida). % accession_number, name keyword(p25195,'plasmid'). % accession_number, string

Related Papers

• Gaurav Marwah and Lois C. Boggess. Artificial Immune Systems for Classification : Some Issues. Department of Computer Science Mississippi State University. [link]
• Prototype Selection for Composite Nearest Neighbor Classifiers. Department of Computer Science University of Massachusetts. 1997. [link]
• Paul Horton and Kenta Nakai. Better Prediction of Protein Cellular Localization Sites with the it k Nearest Neighbors Classifier. ISMB. 1997. [link]
• Andrew Watkins and Jon Timmis and Lois C. Boggess. Artificial Immune Recognition System (AIRS): An ImmuneInspired Supervised Learning Algorithm. (abw5,jt6@kent.ac.uk) Computing Laboratory, University of Kent. [link]
• Mark A. Hall. Department of Computer Science Hamilton, NewZealand Correlation-based Feature Selection for Machine Learning. Doctor of Philosophy at The University of Waikato. 1999. [link]
• Mukund Deshpande and George Karypis. Evaluation of Techniques for Classifying Biological Sequences. PAKDD. 2002. [link]
• Aik Choon Tan and David Gilbert. An Empirical Comparison of Supervised Machine Learning Techniques in Bioinformatics. APBC. 2003. [link]
• [link]