Bottom - Previous - Next - CUBIC - PP home

Explanations for terms (CUBIC's genome analysis)

This page tries to explain some of the jargon used on the CUBIC WWW pages (http://cubic.bioc.columbia.edu).

Please feel free to send questions and comments to the CUBIC administrator: cubic@cubic.bioc.columbia.edu

Contents

General terms used
Transmembrane helix predictions
1. Prediction methods
2. Notations used

General terms used

ORF: open reading frame, typically referring to an expressed gene, resp. a protein.
:

Transmembrane helix prediction for all proteins in entire genome

PHDhtm is a method for the refined prediction of transmembrane helix location and topology. Here, the method was applied to search all proteins from human in Swiss-PROT release 34.0 for transmembrane helices. The results are described in Rost et al., 1996 ( Abstract; Appendix to paper )

Expected accuracy of predictions by PHDhtm

PHDtopology was estimated to have an expected accuracy in predicting all transmembrane helices and the topology correctly of 86% (+/- 3%, one standard deviation). However, for prokaryotes expected accuracy drops to below 73% (+/- 9%, one standard deviation). The expected rate of false positives (proteins without transmembrane helices for which HTM's are falsely predicted) was below 2%; and the expected percentage of proteins with HTM's that were missed was below 3% ( more details ).

Notation used in documents

Notations used in general:

htm: transmembrane helix
location: where in the sequence the transmembrane helix starts and ends
topology: orientation of N-term with respect to membrane
= in if first loop is inside, = out, otherwise
:

Notations used in tables:

FTP : find all data at the CUBIC ftp site : dodo.bioc.columbia.edu in the directory pub/cubic/data/genomes/
HTML : find the respective HTML pages
ASCII : find the respective ASCII pages (load faster than HTML)

ali : alignment of the query protein against proteins in the Swissprot database
phd: secondary structure and solute accessibility prediction of the query protein made by phd
phdHtm: transmembrane helices prediction of the query proteins made by phd
id : sequence identifier
nhtm : number of transmembrane helices
top : topology (in-> N-term, i.e. first non transmembrane region, intra-cytoplasmic, out-> N-term extra-cytoplasmic)
nali : number of sequences in proteinfamily used for prediction
htmCN: positions of predicted transmembrane helices (with respect to first residue)
len1 : length of protein
riTop: reliability of topology prediction (9=high, 0=low)
riMod: reliability of best model (9=high, 0=low)
topD : difference in number of charged loop residues (K+R) = all even loops j- all odd loops
htmCN: position (C-N) of predicted TM helices
seq : sequence of protein
Note: see the ftp version of the entire sequences for details!
seqN : N-term sequence of protein (first 20 residues)
Note: see the ftp version of the entire sequences for details!
seqC : C-term sequence of protein (last 20 residues)
Note: see the ftp version of the entire sequences for details!

Top - Previous - Next - CUBIC - PP home