BENCHMARK OF MEMBRANE HELIX PREDICTIONS FROM SEQUENCE 1



INTRODUCTION :

This benchmark server allows you to compare membrane helix prediction and topology prediction methods that have made their predictions based on amino acid sequence.

If you want to compare methods :

     1.   Choose the methods you want to compare and/or upload your own predictions
     2.   Select the benchmark dataset (standard or personalized)
      3.   Specify the benchmark criteria
      4.   Press 'Execute'


1.   Choose the methods you want to compare and/or upload your own predictions:
DAS1997 (loose) MEMSAT3 PRODIV-TMHMM (in TOPCONS) TMAP Eisenberg (7,10)
DAS1997 (strict) MEMSAT (in TOPCONS-single) PRO-TMHMM (in TOPCONS) TMHMM2 Eisenberg (11,10)
DAS2002 MINNOU SCAMPI TMLOOP Eisenberg (19,10)
DAS-TMfilter OCTOPUS SCAMPI-multi (in TOPCONS) TMMOD Kyte-Doolittle (7,10)
deltaG OCTOPUS (in TOPCONS) SCAMPI-sequence (in TOPCONS) TMPRED Kyte-Doolittle (11,10)
ENSEMBLE (in MemPype) PHDhtm (at PBIL) SCAMPI-sequence (in TOPCONS-single) TOPCONS Kyte-Doolittle (19,10)
HMM-TM PHDThtm (at PBIL) SOSUI TOPCONS-single OHM (7,10)
HMMTOP2 Philius SPLIT4 TOPPRED2 OHM (11,10)
HMMTOP (in TOPCONS-single) Phobius S-TMHMM (in TOPCONS-single) VALPRED OHM (19,10)
MemBrain PolyPhobius SVMtm VALPRED2
MEMSAT-SVM PRED-TMR SVMtop waveTM
If displaying sequences and predictions aligned by residue, choose benchmark standards to display : click to show
If displaying sequences and predictions aligned by residue, choose benchmark standards to display : click to hide
OPM adjusted
show benchmark topology		 OPM
show benchmark topology		 PDBTM
show assigned benchmark topology		 DSSP (shows helices in 3D protein structures)
STRIDE (shows helices in 3D protein structures)
If you have your own membrane helix prediction method that you would like to benchmark against other methods,   click to close
Your results will appear with the method name of YOU
Upload a file containing your predictions :



     or

Enter your predictions in this text box :

The sequences available for benchmarking were derived from 3D protein structure files (PDB files).
The sequence-id xxxx_Y stands for PDB-ID (xxxx) and chain (Y).
Please ensure that your predictions are in one of the following formats :
>3din_E
ooooooooooMMMMMMMMMMMMMMMMMMMMiiiiiiiiiiiiiiiiiiiiiiMMMMMMMMMMMMMMMMMMMooooo
>3h8d_H
>3arc_K
ooooooooooooooooMMMMMMMMMMMMMMMMMMMMMMMMMMiiii
     or

>3din_E
__________MMMMMMMMMMMMMMMMMMMM______________________MMMMMMMMMMMMMMMMMMM_____
>3h8d_H
>3arc_K
________________MMMMMMMMMMMMMMMMMMMMMMMMMM____
     or

>3din_E
11-30,53-71
>3h8d_H
>3arc_K
17-42


2.   Select the benchmark dataset (standard or personalized) :
TRANSMEMBRANE HELIX (TMH) PROTEIN SEQUENCES :
Similarity level : similarity <
Click to hide options If you choose options below such as kingdom, family, or number or type of helices,
you probably want to also set the similarity to 100% so as not to restrict your chosen sequences by similarity.
as measured by :


with similarity measured by :





The use of EMBOSS 'water' (local sequence alignment) is more conservative
than EMBOSS 'needle' (global sequence alignment) in determining
how dissimilar 2 sequences are because proteins having similar conserved domains
inside dissimilar sequences will be judged as having higher similarity
by EMBOSS 'water' than by EMBOSS 'needle'.
Kingdom :
all
eukaryotes only
bacteria only
archaea only
viruses only
none (don't test any TMH sequences)
 Number of TMH :
all
polytopic only
bitopic only
 Type of membrane helices - only sequences having :
all
transmembrane helices only,
          no half-membrane helices or helical reentrant loops
one or more half-membrane helices or helical reentrant loops

Membrane protein structure family : (as in Membrane Proteins of Known 3D Stucture database10)

Restrict to sequences from 3D structures of following year onwards :
not having any homologues prior to this year.
Similarity level of < 30% identity in PDB BLAST with E-value < 0.005.
To perform a blind test using post-2008 data that would not have been used to train most of the prediction methods
choose year 2009.


Choose resolution and experimental method of benchmark data :


Limit benchmark data to models having resolution Å or less.
 X-RAY DIFFRACTION
SOLUTION NMR
ELECTRON CRYSTALLOGRAPHY
ELECTRON MICROSCOPY
FIBER DIFFRACTION
SOLID-STATE NMR
Test for false positives - choose non-transmembrane helix sequences for this benchmark :
MEMBRANE BETA-BARREL PROTEIN SEQUENCES : click to show options
MEMBRANE BETA-BARREL PROTEIN SEQUENCES : click to hide options
Kingdom :
all
eukaryotes only
bacteria only
archaea only
viruses only
none (don't test any membrane beta-barrel sequences)
Similarity level : similarity <

as measured by :


with similarity measured by :
SOLUBLE PROTEIN SEQUENCES : click to show options
SOLUBLE PROTEIN SEQUENCES : click to hide options
Kingdom :
all
eukaryotes only
bacteria only
archaea only
viruses only (include membrane targeting peptides)
none (don't test any membrane soluble protein sequences)
 Number of soluble sequences to include : (or enter a number)

The soluble protein sequences are not similar to each other. They were chosen from PDBselect2514 (< 25% similarity), culled by psi-cd-hit19 to < 1% similarity, divided into kingdoms,
and the top 10% of sequences least similar (by blastall20) to the transmembrane helix (TMH) sequences above were retained and have 17% to 27% identity to the TMH sequences.
Or enter the list of sequences for this benchmark run : click to show options
Enter the list of sequences for this benchmark run : click to hide options


Enter one line per sequence, in the format nnnn_X, where nnnn is the PDB-ID, and X is the chain identifier, eg. 2wll_A
Make sure the sequence ids exist in the benchmark database.
(You can see all valid sequence ids by setting transmembrane helix similarity to 100% and clicking the 'list sequence IDs' option next to the 'Execute' button below.)


3.   Specify the benchmark criteria :
Choose the benchmark standard :
OPM adjusted membrane helices    for benchmarking transmembrane helices and all membrane reentrant loop helices
         (includes all membrane helices that do and don't completely traverse the membrane;
          visual model comparison with proteins of the same family was carried out to assign short helical membrane segments as being membrane helices;
          membrane position relative to each protein is as defined in OPM)
OPM membrane helices6,7    for benchmarking transmembrane helices and some membrane reentrant loop helices longer than a minimum length
         (includes membrane helices that do and some that don't completely traverse the membrane;
          whether a short helical membrane segment is classified in OPM as a membrane helix depends on length of helix and depth into membrane;
          membrane position relative to each protein is as defined in OPM)
PDBTM membrane helices and loops8,9    for benchmarking transmembrane helices and all membrane reentrant loop helices and their associated non-helical coil
         (includes membrane helices that do and don't completely traverse the membrane, and their associated membrane coil forming reentrant hairpin-like loop;
          membrane coil and short helical membrane segments are classified here as a membrane loop if they are part of a reentrant loop leaving the PDBTM-defined membrane on the same side they entered;
          membrane position relative to each protein is as defined in PDBTM)
PDBTM membrane helices8,9    for benchmarking transmembrane helices and not any membrane helices belonging to reentrant loops
         (includes membrane helices that do traverse the membrane, and excludes those that don't completely traverse the membrane;
          membrane position relative to each protein is as defined in PDBTM)


Adjust the benchmark parameters :
Minimum length of observed and predicted helices (smaller helices are ignored)  : 
Minimum overlap of observed helix residues for a helix prediction to score in the
per-sequence-accuracy, per-segment-accuracy and average-helix-boundary-difference scores		  : 
Maximum distance in residues to include in the membrane-helix-boundary scores  : 

Sort by benchmark results on :

Graph the benchmark results for :






Graph title :   


Parameters for viewing sequences and predictions aligned by residue :
Display width in residues (break to a new line after this number of residues, zero means don't break at all) :

OPM membrane helices are defined at Orientations of Proteins in Membranes (OPM) database
PDBTM membrane helices and loops are defined at PDBTM: Protein Data Bank of Transmembrane Proteins

click for less criteria


4.   Press 'Execute' :



















Run a benchmark for chosen sequences and prediction methods :

run the benchmark (may take a few seconds to run)

Display chosen sequences and prediction methods :

display sequences and predictions aligned by residue
display sequences and predictions aligned by residue, and other information

Display chosen sequences :

display FASTA sequences (missing/shifted residues appear as _ )
display FASTA sequences (missing/shifted residues appear as G for glycine)
download FASTA sequences (missing/shifted residues appear as _ )
download FASTA sequences (missing/shifted residues appear as G for glycine)
list sequence IDs
list sequence IDs, names and other information
count the number of sequences

The data for the benchmarking standard is from protein models in the Protein Data Bank (PDB)11.

FASTA sequence is from the corresponding PDB FASTA file and sometimes contains more residues than appear in the PDB model file.

Where necessary, FASTA sequences have been shifted to conserve the residue numbering in the PDB model file.

Missing residues and sequence positions of shifted residues were set to G for Glycine when submitted to prediction programs for this benchmark.


KEY TO BENCHMARK RESULTS : click to show
KEY TO BENCHMARK RESULTS : click to hide


TOPOGRAPHY SCORES (for scoring predictions of membrane helices versus not membrane helix) :


Per protein sequence accuracy score :
  • Qok% : Percentage of protein sequences for which all membrane helices are predicted correctly
 
Qok%    =      		number of protein sequences (chains) having
all observed membrane helices are predicted by the prediction method
and all predicted membrane helices are actually observed in the benchmark data
______________________________________________________________________   *   100

number of protein sequences (chains)
Per segment accuracy scores :
  • Qhtm %obs : Percentage of all observed membrane helices that are predicted correctly
       (sensitivity)
 
Qhtm %obs    =      		number of membrane helices observed in the benchmark data
that the prediction method did predict as being membrane helices
______________________________________________________________________   *   100

number of membrane helices observed in the benchmark data
  • Qhtm %prd : Percentage of all predicted membrane helices that are predicted correctly
       (specificity)
 
Qhtm %prd    =      		number of membrane helices predicted by the prediction method
that are actually observed in the benchmark data as being membrane helices
______________________________________________________________________   *   100

number of membrane helices predicted by the prediction method
Helix boundary accuracy scores :
  • AvHb diff : Average helix boundary position difference in residues for the prediction versus the observed helix boundaries
 
AvHb diff    =      		∑ | ( distance in residues of observed helix boundary minus predicted helix boundary ) |
____________________________________________________________

number of membrane helices observed in the benchmark data          *     2
that the prediction method did predict as being membrane helices
 
( non-predicted observed helices and non-observed predicted helices are not included )
  • QHb %obs : Percentage of all observed membrane helix boundaries (2 per helix) that are predicted correctly (within distance from observed boundary)
 
QHb %obs    =      		number of membrane helix boundaries observed in the benchmark data
for which the prediction method predicted the helix boundary
within a certain number of residues from the boundary actually observed in the benchmark data
______________________________________________________________________   *   100

number of membrane helices observed in the benchmark data * 2
  • Gauss QHb %obs : Scaled percentage of all observed membrane helix boundaries (2 per helix) that are predicted correctly
    (score is scaled as a Gaussian curve normal distribution (mean = 0, variance = std-dev2 = 5 or chosen by user) around the observed helix boundary because observed helix boundary may not be exact)
 
Gauss QHb %obs    =      		∑ ( score for each membrane helix boundary predicted by the prediction method )
____________________________________________________________   *   100

number of membrane helices observed in the benchmark data * 2
 
standard-deviation    =     Maximum distance in residues to include in the membrane-helix-boundary scores
 
distance    =     predicted helix boundary - observed helix boundary
 
score for each membrane helix boundary    =     
predicted by the prediction method
      score-numerator      
    score-denominator    
( if predicted helix boundary is within or equal to 2 * standard-deviation from the observed helix boundary;
  otherwise score = 0 )
 
score-numerator    =     
                         1                         
standard-deviation * √ ( 2 * Π )
   *   e 



 		- ( distance - mean )2 / (2 * standard-deviation2 ) )
 
score-denominator    =     
                         1                         
standard-deviation * √ ( 2 * Π )
   *   e 



 		- ( mean )2 / (2 * standard-deviation2 ) )
     =     
                         1                         
standard-deviation * √ ( 2 * Π )
  ( Thus, when predicted helix boundary is same as observed helix boundary, then distance = 0, and score = 1 or 100% )
Per residue accuracy scores :
  • Q2% : Percentage of correctly predicted residues in two-states: membrane helix / not part of a membrane helix
                    (each sequence contributes equally so that long sequences don't dominate the score)
 
Q2%    =      		              number of residues in a sequence that were correctly predicted as being membrane helix or not
(   ∑   ( ________________________________________________________________________________ )   )   *   100   /   number of sequences

                                                      number of residues in that sequence
  • htm MCC : Matthews Correlation Co-efficient* for prediction of residues as membrane helical vs. not membrane helical
 
htm MCC    =      		                         (TP*TN) - (FP*FN)
________________________________________

     √ ( (TP+FP) * (TP+FN) * (TN+FP) * (TN+FN) )

TP (true positives) = number of correctly predicted membrane helix residues
TN (true negatives) = number of residues correctly predicted as not in a membrane helix
FP (false positives) = number of incorrectly predicted membrane helix residues
FN (false negatives) = number of residues incorrectly predicted as not in a membrane helix
* Matthews B (1975) Comparison of the predicted and observed secondary structure of T4 phage lysozyme. Biochimica et Biophysica acta, 405, 442-451.
  • Q2T %obs : Percentage of all observed membrane helix residues that are predicted correctly
 
Q2T %obs    =      		number of residues predicted as being part of a membrane helix
that really are observed in the benchmark data as being in a membrane helix
______________________________________________________________________   *   100

number of residues that are observed in the benchmark data as being in a membrane helix
  • Q2T %prd : Percentage of all predicted membrane helix residues that are predicted correctly
 
Q2T %prd    =      		number of residues predicted as being part of a membrane helix
that really are observed in the benchmark data as being in a membrane helix
______________________________________________________________________   *   100

number of residues predicted as being part of a membrane helix
  • Q2N %obs : Percentage of all observed non-membrane helix residues that are predicted correctly
 
Q2N %obs    =      		number of residues predicted as NOT being part of a membrane helix
that really are observed in the benchmark data as NOT being in a membrane helix
______________________________________________________________________   *   100

number of residues that are observed in the benchmark data as NOT being in a membrane helix
  • Q2N %prd : Percentage of all predicted non-membrane helix residues that are predicted correctly
 
Q2N %prd    =      		number of residues predicted as NOT being part of a membrane helix
that really are observed in the benchmark data as NOT being in a membrane helix
______________________________________________________________________   *   100

number of residues predicted as NOT being in a membrane helix


TOPOLOGY SCORES (for scoring predictions of inside/outside topology) :


Per protein sequence accuracy score :
  • Qok3% : Percentage of protein sequences for which all inside, outside and membrane helix topologies are predicted correctly
 
Qok3%    =      		number of protein sequences (chains) having
all observed segment topologies (inside, outside or membrane helix) are predicted by the prediction method
and all predicted segment topologies are actually observed in the benchmark data
__________________________________________________________________________________________   *   100

number of protein sequences (chains)
  • Nterm % : Percentage of protein sequences for which the topology of the N-terminal topology is predicted correctly
 
Nterm %    =      		number of protein sequences (chains) having
N-terminal topology predicted correctly (inside or outside side of membrane)
__________________________________________________________________________________________   *   100

number of protein sequences (chains)
Per segment accuracy scores :
  • ioSeg Q2% : Percentage of correctly predicted topology segments in two-states: inside side / outside side of membrane
                                (each sequence contributes equally so that a topology prediction that starts correctly and finishes incorrectly
                                due to a missed membrane segment prediction rather than due to an incorrect topology prediction start
                                will be penalised as an incompletely correct topology prediction (or not completely incorrect prediction) and long sequences will not be overly penalised in the final score)
 
ioSeg Q2%    =      		            number of correctly predicted non-membrane topologies
            (inside side or outside side of membrane) predicted in a sequence
(   ∑   ( _____________________________________________________________________________________ )   )   *   100   /   number of sequences

            number of non-membrane topologies (inside side or outside side of membrane) observed in that sequence
  • Qiom %obs : Percentage of all observed topologies (inside, outside or membrane helix) that are predicted correctly
 
Qiom %obs    =      		number of topology segments (inside, outside or membrane helix) observed in the benchmark data
that the prediction method did predict as being the correct type of topology
__________________________________________________________________________________________   *   100

number of topology segments observed in the benchmark data
  • Qio %obs : Percentage of all observed non-membrane topologies (inside side or outside side of membrane) that are predicted correctly
 
Qio %obs    =      		number of non-membrane topologies (inside side or outside side of membrane) observed in the benchmark data
that the prediction method did predict as being the correct type of non-membrane topology
__________________________________________________________________________________________   *   100

number of non-membrane topologies observed in the benchmark data
Per residue accuracy scores :
  • Q3% : Percentage of correctly predicted residues in three-states: membrane helix / inside non-membrane residue / outside non-membrane residue
                    (each sequence contributes equally so that long sequences don't dominate the score)
 
Q3%    =      		              number of residues in a sequence that were correctly predicted as being
              membrane helix, inside non-membrane residue or outside non-membrane residue
(   ∑   ( ______________________________________________________________________ )   )   *   100   /   number of sequences

                                                      number of residues in that sequence
  • ioRes Q2% : Percentage of correctly predicted residues in two-states: inside non-membrane residue / outside non-membrane residue
                                (each sequence contributes equally so that long sequences don't dominate the score)
 
ioRes Q2%    =      		              number of non-membrane residues in a sequence that were correctly predicted as being
              inside non-membrane or outside non-membrane residue
(   ∑   ( ______________________________________________________________________ )   )   *   100   /   number of sequences

                                                      number of non-membrane residues in that sequence
  • io MCC : Matthews Correlation Co-efficient* for topology prediction of residues as either on the inside vs. the outside side of the membrane
 
io MCC    =      		                         (TP*TN) - (FP*FN)
________________________________________

     √ ( (TP+FP) * (TP+FN) * (TN+FP) * (TN+FN) )

TP (true positives) = number of non-membrane residues on the inside side of the membrane that are correctly predicted
TN (true negatives) = number of non-membrane residues on the inside side of the membrane
FP (false positives) = number of non-membrane residues on the outside side of the membrane that are correctly predicted
FN (false negatives) = number of non-membrane residues on the outside side of the membrane
* Matthews B (1975) Comparison of the predicted and observed secondary structure of T4 phage lysozyme. Biochimica et Biophysica acta, 405, 442-451.

Inside refers to the inside side of the membrane, and outside refers to the outside side of the membrane.


KEY TO DISPLAY OF ALIGNED SEQUENCES AND PREDICTIONS : click to show
KEY TO DISPLAY OF ALIGNED SEQUENCES AND PREDICTIONS : click to hide

OPM6,7 :

M : transmembrane secondary structure segment i : inside (on the inside side of the membrane)
   (may be α-helix for membrane helical proteins o : outside (on the outside side of the membrane)
    or β-sheet for beta-barrel membrane proteins) _ : residue position unknown - not in PDB file
m : membrane residue not part of transmembrane segment
   (residue's α-carbon used to decide if residue in membrane or not)

PDBTM8,9 :

H : alpha helix L : membrane loop 1 : side 1 C : coil i : inside (assigned by comparing PDBTM sides 1 and 2 to OPM)
B : beta strand I : membrane inside 2 : side 2 U : unknown o : outside (assigned by comparing PDBTM sides 1 and 2 to OPM)

DSSP12 :

H : α helix E : β sheet T : turn
I : π helix B : β bridge S : high curvature region
G : 310 helix _ : other (loop) or unknown

STRIDE13 :

H : α helix E      : Extended conformation T : turn
I : π helix B or b : β isolated bridge S : high curvature region
G : 310 helix C      : coil (none of the above) _ : residue position unknown - not in PDB file

ENSEMBLE in (MemPype), HMM-TM, HMMTOP (in TOPCONS-single), MEMSAT (in TOPCONS-single), OCTOPUS (in TOPCONS), proT, prodivT,
SCAMPI, SCAMPI-multi (multiple sequence alignment) (in TOPCONS), SCAMPI-sequence (in TOPCONS-single), SCAMPI-sequence (in TOPCONS),
TMHMM2, S-TMHMM (in TOPCONS-single), TMMOD, TOPCONS, TOPCONS-single :

M : membrane α helix i : inside (on the inside side of the membrane)
o : outside (on the outside side of the membrane)

SVMtop :

H : membrane α helix i : inside (on the inside side of the membrane)
o : outside (on the outside side of the membrane)

TOPPRED2 :

M : membrane α helix i : inside (on the inside side of the membrane)
l : non-membrane loop o : outside (on the outside side of the membrane)

DAS2002, DAS1997, deltaG, MemBrain, MINNOU, PRED-TMR, SOSUI, SPLIT4, SVMtm, TMAP, VALPRED, VALPRED2, waveTM,
Eisenberg (7,10), Eisenberg (11,10), Eisenberg (19,10), Kyte-Doolittle (7,10), Kyte-Doolittle (11,10), Kyte-Doolittle (19,10), OHM (7,10), OHM (11,10), OHM (19,10) :

M : membrane α helix _ : not a membrane α helix

PHDhtm (at PBIL) :

H : membrane α helix _ : not a membrane α helix

PHDThtm (at PBIL) :

T : transmembrane region i : intra-cytoplasmic
_ : not a membrane α helix o : extra-cytoplasmic

DAS-TMfilter :

M : membrane α helix S : Potential signal peptide _ : not a membrane α helix

OCTOPUS :

M : the hydrophobic part of the membrane, 0-13Å from the membrane center i : inside side of the membrane
L : a close loop region, 13-23Å from the membrane center o : outside side of the membrane
R : reentrant g : a globular region, further than 23Å from the membrane
I : the membrane water-interface, 11-18Å from the membrane center

HMMTOP2 :

H : membrane helix i : inside helix tail (not in membrane) I : inside loop
o : outside helix tail (not in membrane) O : outside loop

MEMSAT-SVM, Philius, Phobius, PolyPhobius :

H : membrane helix i : inside side of the membrane
S : signal peptide o : outside side of the membrane

MEMSAT3 :

H : Central transmembrane helix segment S : Possible N-terminal signal peptide
I : Inside helix cap + : Inside loop
O : Outside helix cap - : Outside loop

TMLOOP :

M : membrane α helix _ : not a membrane α helix



REFERENCES USED IN CREATING BENCHMARK SERVER : click to show
REFERENCES USED IN CREATING BENCHMARK SERVER : click to hide
[1] Emma M. Rath, Dominique Tessier, Alexander A. Campbell, Hong Ching Lee, Tim Werner, Noeris K. Salam, Lawrence K. Lee, W. Bret Church (2013)
'A benchmark server using high resolution protein structure data, and benchmark results for membrane helix predictions'
BMC Bioinformatics, 14, 111 doi: 10.1186/1471-2105-14-111. PMID 23530628
[2] Lawrence K. Lee, Noeris K. Salam, and W. Bret Church (in preparation)
'Transmembrane helix analysis using threading approaches'
[3] Kernytsky A, Rost B (2003)
'Static benchmarking of membrane helix predictions'
Nucleic Acids Research, 31, 3642-3644. PMID 12824384
[4] Chen CP, Kernytsky A, Rost B (2002)
'Transmembrane helix predictions revisited'
Protein Science, 11, 2774-2791. PMID 12441377
[5] Zemla A, Venclovas Č, Fidelis K, Rost B (1999)
'A Modified Definition of Sov, a Segment-Based Measure for Protein Secondary Structure Prediction Assessment'
PROTEINS: Structure, Function, and Genetics, 34, 220-223. PMID 10022357
[6] Lomize MA, Lomize AL, Pogozheva ID, Mosberg HI (2006)
'OPM: Orientations of Proteins in Membranes database'
Bioinformatics, 22, 623-625. PMID 16397007
[7] Lomize MA, Pogozheva ID, Joo H, Mosberg HI, Lomize AL (2012)
OPM database and PPM web server: resources for positioning of proteins in membranes.
Nucleic Acids Research, 40(Database issue), D370-376. PMID 21890895
[8] Tusnády GE, Dosztányi Z, Simon I. (2004)
'Transmembrane proteins in the Protein Data Bank: identification and classification.'
Bioinformatics, 20, 2964-2972. PMID 15180935
[9] Tusnády GE, Dosztányi Z, Simon I. (2005)
'PDB_TM: selection and membrane localization of transmembrane proteins in the protein data bank.'
Nucleic Acids Research, 33(Database issue), D275-8. PMID 15608195
[10] White SH. (2009)
Biophysical dissection of membrane proteins.
Nature, 459, 344-346. PMID 19458709
[11] Berman HM, Westbrook J, Feng Z, Gilliland G, Bhat TN, Weissig H, Shindyalov IN, Bourne PE. (2000)
The Protein Data Bank.
Nucleic Acids Research, 28, 235-242. PMID 10592235
[12] Kabsch W, Sander C (1983)
Dictionary of protein secondary structure: pattern recognition of hydrogen-bonded and geometrical features.
Biopolymers, 22, 2577-637. PMID 6667333
[13] Frishman D, Argos P (1995)
Knowledge-based secondary structure assignment.
Proteins: structure, function and genetics, 23, 566-579. PMID 8749853
[14] Hobohm U, Scharf M, Schneider R, Sander C (1992)
Selection of representative protein data sets.
Protein Science, 1, 409-417. PMID 1304348
[15] Sander C, Schneider R (1991)
Database of Homology-Derived Protein Structures and the Structural Meaning of Sequence Alignment.
Proteins: Structure, Function, and Genetics, 9, 56-68. PMID 2017436
[16] Needleman SB, Wunsch CD (1970)
A general method applicable to the search for similarities in the amino acid sequence of two proteins.
Journal of Molecular Biology, 48, 443-453. PMID 5420325
[17] Smith TF, Waterman MS (1981)
Identification of Common Molecular Subsequences.
Journal of Molecular Biology, 147, 195-197. PMID 7265238
[18] Rice P, Longden I, Bleasby A (2000)
EMBOSS: the European Molecular Biology open software suite.
Trends in Genetics, 16, 276-277. PMID 10827456
[19] Li W, Godzik A (2006)
Cd-hit: a fast program for clustering and comparing large sets of protein or nucleotide sequences.
Bioinformatics, 22, 13, 1658-1659. PMID 16731699
[20] Altschul S, Gish W, Miller W, Myers E, Lipman D (1990)
Basic local alignment search tool.
Journal of Molecular Biology, 215, 3, 403-410. PMID 2231712


METHODS BENCHMARKED IN BENCHMARK SERVER : click to show
METHODS BENCHMARKED IN BENCHMARK SERVER : click to hide
DASTMfilt DAS-TMfilter
http://www.enzim.hu/DAS/DAS.html
Cserzo M, Eisenhaber F, Eisenhaber B, Simon I (2004)
TM or not TM: transmembrane protein prediction with low false positive rate using DAS-TMfilter. Bioinformatics, 20, 1, 136-137. PMID 14693825
(uses sequence alignment of membrane protein sequences and hydropathy dot-plots)
DAS2002 DAS2002
http://mendel.imp.ac.at/DAS/
Cserzo M, Eisenhaber F, Eisenhaber B, Simon I (2002)
On filtering false positive transmembrane protein predictions. Protein Engineering, 15, 745-752. PMID 12456873
(uses sequence alignment of membrane protein sequences and hydropathy dot-plots)
DAS1997l DAS1997 (loose)
http://www.sbc.su.se/~miklos/DAS/ cutoff=1.7
Cserzo M, Wallin E, Simon I, von Heijne G, Elofsson A (1997)
Prediction of transmembrane alpha-helices in procariotic membrane proteins: the Dense Alignment Surface method. Protein Engineering, 10, 673-676. PMID 9278280
(uses sequence alignment of membrane protein sequences and hydropathy dot-plots)
DAS1997s DAS1997 (strict)
http://www.sbc.su.se/~miklos/DAS/ cutoff=2.2
Cserzo M, Wallin E, Simon I, von Heijne G, Elofsson A (1997)
Prediction of transmembrane alpha-helices in procariotic membrane proteins: the Dense Alignment Surface method. Protein Engineering, 10, 673-676. PMID 9278280
(uses sequence alignment of membrane protein sequences and hydropathy dot-plots)
deltaG deltaG
http://www.cbr.su.se/DGpred/
Hessa T, Meindl-Beinker N, Bernsel A, Kim J, Sato Y, Lerch M, Lundin C, Nilsson I, White SH, von Heijne G (2007)
Molecular code for transmembrane-helix recognition by the Sec61 translocon. Nature, 450, 1026-1030. PMID 18075582
(uses experimentally derived biophysical residue free energy values for insertion into membranes)
Eisen(7) Eisenberg (7,10)
ftp://emboss.open-bio.org/pub/EMBOSS/EMBOSS-6.4.0.tar.gz EMBOSS 6.4.0 pepinfo method=Eisenberg window=7 cutoff=10, minimum helix length of 10
Eisenberg D, Weiss RM, Terwilliger TC (1982)
The helical hydrophobic moment: a measure of the amphiphilicity of a helix. Nature, 299, 371-374. PMID 7110359
Rice P, Longden I, Bleasby A (2000)
EMBOSS: the European Molecular Biology open software suite. Trends in Genetics, 16, 276-277. PMID 10827456
(uses biophysical residue hydropathy values)
Eisen(11) Eisenberg (7,10)
ftp://emboss.open-bio.org/pub/EMBOSS/EMBOSS-6.4.0.tar.gz EMBOSS 6.4.0 pepinfo method=Eisenberg window=11 cutoff=10, minimum helix length of 10
Eisenberg D, Weiss RM, Terwilliger TC (1982)
The helical hydrophobic moment: a measure of the amphiphilicity of a helix. Nature, 299, 371-374. PMID 7110359
Rice P, Longden I, Bleasby A (2000)
EMBOSS: the European Molecular Biology open software suite. Trends in Genetics, 16, 276-277. PMID 10827456
(uses biophysical residue hydropathy values)
Eisen(19) Eisenberg (7,10)
ftp://emboss.open-bio.org/pub/EMBOSS/EMBOSS-6.4.0.tar.gz EMBOSS 6.4.0 pepinfo method=Eisenberg window=19 cutoff=10, minimum helix length of 10
Eisenberg D, Weiss RM, Terwilliger TC (1982)
The helical hydrophobic moment: a measure of the amphiphilicity of a helix. Nature, 299, 371-374. PMID 7110359
Rice P, Longden I, Bleasby A (2000)
EMBOSS: the European Molecular Biology open software suite. Trends in Genetics, 16, 276-277. PMID 10827456
(uses biophysical residue hydropathy values)
ENSEMBLE ENSEMBLE (in the MemPype server)
http://mu2py.biocomp.unibo.it/mempype
Martelli PL, Fariselli P, Casadio R. (2003)
An ENSEMBLE machine learning approach for the prediction of all-alpha membrane proteins. Bioinformatics, 19, Suppl 1, i205-11. PMID 12855459
Pierleoni A, Indio V, Savojardo C, Fariselli P, Martelli PL, Casadio R. (2011)
MemPype: a pipeline for the annotation of eukaryotic membrane proteins. Nucleic Acids Res, 39(Web Server issue), W375-80. PMID 21543452
(uses neural network (NN) and hidden Markov models (HMM) trained on membrane protein sequences)
HMM-TM HMM-TM
http://bioinformatics.biol.uoa.gr/HMM-TM/
Bagos PG, Liakopoulos TD, Hamodrakas SJ (2006)
Algorithms for incorporating prior topological information in HMMs: application to transmembrane proteins. BMC Bioinformatics, 5, 7, 189. PMID 16597327
(uses hidden Markov model (HMM) trained on membrane protein sequences)
HMMTOP2 HMMTOP2
http://www.enzim.hu/hmmtop/
Tusnády GE, Simon I (1998)
Principles governing amino acid composition of integral membrane proteins: applications to topology prediction. Journal of Molecular Biology, 283, 489-506. PMID 9769220
Tusnády GE, Simon I (2001)
The HMMTOP transmembrane topology prediction server. Bioinformatics, 17, 849-850. PMID 11590105
(uses hidden Markov model (HMM) trained on membrane protein sequences)
hmmtopS HMMTOP (in the TOPCONS-single server)
http://single.topcons.net
Tusnády GE, Simon I (1998)
Principles governing amino acid composition of integral membrane proteins: applications to topology prediction. Journal of Molecular Biology, 283, 489-506. PMID 9769220
Tusnády GE, Simon I (2001)
The HMMTOP transmembrane topology prediction server. Bioinformatics 17, 849-850. PMID 11590105
Hennerdal A, Elofsson A (2011)
Rapid membrane protein topology prediction. Bioinformatics 27, 9, 1322-1323. PMID 21493661
(uses hidden Markov model (HMM) trained on membrane protein sequences)
KyteD(7) Kyte-Doolittle (7,10)
ftp://emboss.open-bio.org/pub/EMBOSS/EMBOSS-6.4.0.tar.gz EMBOSS 6.4.0 pepinfo method=Kyte-Doolittle window=7 cutoff=10, minimum helix length of 10
Kyte J, Doolittle RF (1982)
A simple method for displaying the hydropathic character of a protein. Journal of Molecular Biology, 157, 105-132. PMID 7108955
Rice P, Longden I, Bleasby A (2000)
EMBOSS: the European Molecular Biology open software suite. Trends in Genetics, 16, 276-277. PMID 10827456
(uses biophysical residue hydropathy values)
KyteD(11) Kyte-Doolittle (7,10)
ftp://emboss.open-bio.org/pub/EMBOSS/EMBOSS-6.4.0.tar.gz EMBOSS 6.4.0 pepinfo method=Kyte-Doolittle window=11 cutoff=10, minimum helix length of 10
Kyte J, Doolittle RF (1982)
A simple method for displaying the hydropathic character of a protein. Journal of Molecular Biology, 157, 105-132. PMID 7108955
Rice P, Longden I, Bleasby A (2000)
EMBOSS: the European Molecular Biology open software suite. Trends in Genetics, 16, 276-277. PMID 10827456
(uses biophysical residue hydropathy values)
KyteD(19) Kyte-Doolittle (7,10)
ftp://emboss.open-bio.org/pub/EMBOSS/EMBOSS-6.4.0.tar.gz EMBOSS 6.4.0 pepinfo method=Kyte-Doolittle window=19 cutoff=10, minimum helix length of 10
Kyte J, Doolittle RF (1982)
A simple method for displaying the hydropathic character of a protein. Journal of Molecular Biology, 157, 105-132. PMID 7108955
Rice P, Longden I, Bleasby A (2000)
EMBOSS: the European Molecular Biology open software suite. Trends in Genetics, 16, 276-277. PMID 10827456
(uses biophysical residue hydropathy values)
MemBrain MemBrain
http://chou.med.harvard.edu/bioinf/MemBrain
Shen H, Chou JJ (2008)
MemBrain: improving the accuracy of predicting transmembrane helices. PLoS One. 11, 3, e2399. PMID 18545655
(uses sequence alignment of membrane protein sequences and machine learning trained on membrane protein sequences)
MEMSAT-SVM MEMSAT-SVM
http://bioinf.cs.ucl.ac.uk/psipred/?program=svmmemsat
Nugent T, Jones DT (2009)
Transmembrane protein topology prediction using support vector machines. BMC Bioinformatics, 10, 159. PMID 19470175
(uses support vector machine (SVM) trained on membrane protein sequences)
MEMSAT3 MEMSAT3
http://bioinf.cs.ucl.ac.uk/psipred/?program=svmmemsat
Jones DT (2007)
Improving the accuracy of transmembrane protein topology prediction using evolutionary information. Bioinformatics, 23, 5, 538-544. PMID 17237066
Jones DT, Taylor WR, Thornton JM (1994)
A model recognition approach to the prediction of all-helical membrane protein structure and topology. Biochemistry, 33, 10, 3038-3049. PMID 8130217
(uses artificial neural network (NN) trained on membrane protein sequences)
memsatS MEMSAT (in the TOPCONS-single server)
http://single.topcons.net
Jones DT, Taylor WR, Thornton JM (1994)
A model recognition approach to the prediction of all-helical membrane protein structure and topology. Biochemistry, 33, 3038-3049. PMID 8130217
Hennerdal A, Elofsson A (2011)
Rapid membrane protein topology prediction. Bioinformatics 27, 9, 1322-1323. PMID 21493661
(uses artificial neural network (NN) trained on membrane protein sequences)
minnou MINNOU
http://minnou.cchmc.org/
Cao B, Porollo A, Adamczak R, Jarrell M, Meller J (2006)
Enhanced Recognition of Protein Transmembrane Domains with Prediction-based Structural Profiles. Bioinformatics 22:303-309. PMID 16293670
(uses biophysical residue properties and solvent exposure of predicted secondary structure determined from multiple sequence alignments)
OCTOPUS OCTOPUS
http://octopus.cbr.su.se/
Viklund H, Elofsson A (2008)
OCTOPUS: improving topology prediction by two-track ANN-based preference scores and an extended topological grammar. Bioinformatics, 24, 15, 1662-1668. PMID 18474507
Viklund H, Bernsel A, Skwark M, Elofsson A (2008)
SPOCTOPUS: a combined predictor of signal peptides and membrane protein topology. Bioinformatics, 24, 24, 2928-2929. PMID 18945683
(uses hidden Markov models (HMM) and artificial neural networks (NN) trained on membrane protein sequences)
octopusT OCTOPUS (in the TOPCONS server)
http://topcons.cbr.su.se/
Viklund H, Elofsson A (2008)
OCTOPUS: improving topology prediction by two-track ANN-based preference scores and an extended topological grammar. Bioinformatics, 24, 15, 1662-1668. PMID 18474507
Viklund H, Bernsel A, Skwark M, Elofsson A (2008)
SPOCTOPUS: a combined predictor of signal peptides and membrane protein topology. Bioinformatics, 24, 24, 2928-2929. PMID 18945683
Bernsel A, Viklund H, Hennerdal A, Elofsson A (2009)
TOPCONS: consensus prediction of membrane protein topology. Nucleic Acids Research, Web Server Issue 37, W465-W468. PMID 19429891
(uses hidden Markov models (HMM) and artificial neural networks (NN) trained on membrane protein sequences)
OHM(7) OHM (7,10)
ftp://emboss.open-bio.org/pub/EMBOSS/EMBOSS-6.4.0.tar.gz EMBOSS 6.4.0 pepinfo method=OHM window=7 cutoff=10, minimum helix length of 10
Sweet RM, Eisenberg D (1983)
Correlation of sequence hydrophobicities measures similarity in three-dimensional protein structure. Journal of Molecular Biology, 171, 479-488. PMID 6663622
Rice P, Longden I, Bleasby A (2000)
EMBOSS: the European Molecular Biology open software suite. Trends in Genetics, 16, 276-277. PMID 10827456
(uses biophysical residue hydropathy values)
(uses hidden Markov models (HMM) and artificial neural networks (NN) trained on membrane protein sequences)
OHM(11) OHM (7,10)
ftp://emboss.open-bio.org/pub/EMBOSS/EMBOSS-6.4.0.tar.gz EMBOSS 6.4.0 pepinfo method=OHM window=11 cutoff=10, minimum helix length of 10
Sweet RM, Eisenberg D (1983)
Correlation of sequence hydrophobicities measures similarity in three-dimensional protein structure. Journal of Molecular Biology, 171, 479-488. PMID 6663622
Rice P, Longden I, Bleasby A (2000)
EMBOSS: the European Molecular Biology open software suite. Trends in Genetics, 16, 276-277. PMID 10827456
(uses biophysical residue hydropathy values)
OHM(19) OHM (19,10)
ftp://emboss.open-bio.org/pub/EMBOSS/EMBOSS-6.4.0.tar.gz EMBOSS 6.4.0 pepinfo method=OHM window=19 cutoff=10, minimum helix length of 10
Sweet RM, Eisenberg D (1983)
Correlation of sequence hydrophobicities measures similarity in three-dimensional protein structure. Journal of Molecular Biology, 171, 479-488. PMID 6663622
Rice P, Longden I, Bleasby A (2000)
EMBOSS: the European Molecular Biology open software suite. Trends in Genetics, 16, 276-277. PMID 10827456
(uses biophysical residue hydropathy values)
PHDhtm PHDhtm (at PBIL)
http://npsa-pbil.ibcp.fr/cgi-bin/npsa_automat.pl?page=/NPSA/npsa_htm.html
Rost B, Sander C (1994)
Combining evolutionary information and neural networks to predict protein secondary structure. Proteins, 19, 55-72. PMID 8066087
Rost B, Sander C (1993)
Prediction of protein secondary structure at better than 70% accuracy. J. Mol. Biol., 232, 584-599. PMID 8345525
Rost B, Sander C (1993)
Improved prediction of protein secondary structure by use of sequence profiles and neural networks. Proc. Natl. Acad. Sci. U.S.A., 90, 7558-7562. PMID 8356056
Rost B, Casadio R, Fariselli P, Sander C (1995)
Transmembrane helices predicted at 95% accuracy. Protein Sci, 4, 3, 521-533. PMID 7795533
Combet C, Blanchet C, Geourjon C, Deléage G (2000)
NPS@: Network Protein Sequence Analysis. TIBS 2000, 3, 291, 147-150. PMID 10694887
(uses sequence alignment of membrane protein sequences and artificial neural network (NN) trained on membrane protein sequences)
PHDThtm PHDThtm (at PBIL)
http://npsa-pbil.ibcp.fr/cgi-bin/npsa_automat.pl?page=/NPSA/npsa_htm.html
Rost B, Sander C (1994)
Combining evolutionary information and neural networks to predict protein secondary structure. Proteins, 19, 55-72. PMID 8066087
Rost B, Sander C (1993)
Prediction of protein secondary structure at better than 70% accuracy. J. Mol. Biol., 232, 584-599. PMID 8345525
Rost B, Sander C (1993)
Improved prediction of protein secondary structure by use of sequence profiles and neural networks. Proc. Natl. Acad. Sci. U.S.A., 90, 7558-7562. PMID 8356056
Rost B, Casadio R, Fariselli P, Sander C (1995)
Transmembrane helices predicted at 95% accuracy. Protein Sci, 4, 3, 521-533. PMID 7795533
Combet C, Blanchet C, Geourjon C, Deléage G (2000)
NPS@: Network Protein Sequence Analysis. TIBS 2000, 3, 291, 147-150. PMID 10694887
(uses sequence alignment of membrane protein sequences and artificial neural network (NN) trained on membrane protein sequences)
Philius Philius
http://www.yeastrc.org/philius/pages/philius/uploadFASTA.jsp
http://www.yeastrc.org/philius/pages/philius/runPhilius.jsp
Reynolds SM, Käll L, Riffle ME, Bilmes JA, Noble WS (2008)
Transmembrane topology and signal peptide prediction using dynamic bayesian networks. PLoS Comput Biol, 4, 11, e1000213. PMID 18989393
(uses dynamic Bayesian network trained on membrane protein sequences)
Phobius Phobius
http://phobius.cgb.ki.se
Käll L, Krogh A, Sonnhammer ELL (2004)
A Combined Transmembrane Topology and Signal Peptide Prediction Method. J Mol Biol, 14, 5, 1027-1036. PMID 15111065
Käll L, Krogh A, Sonnhammer EL (2007)
Advantages of combined transmembrane topology and signal peptide prediction--the Phobius web server. Nucleic Acids Res, 35, Web Server issue, W429-432. PMID 17483518
(uses hidden Markov model (HMM) trained on membrane protein sequences)
PolyPhobs PolyPhobius
http://phobius.sbc.su.se/poly.html
Käll L, Krogh A, Sonnhammer ELL (2005)
An HMM posterior decoder for sequence feature prediction that includes homology information. Bioinformatics, 21, Suppl 1, i251-257. PMID 15961464
Käll L, Krogh A, Sonnhammer EL (2007)
Advantages of combined transmembrane topology and signal peptide prediction--the Phobius web server. Nucleic Acids Res, 35, Web Server issue, W429-432. PMID 17483518
(uses hidden Markov model (HMM) trained on membrane protein sequences and sequence alignment of membrane protein sequences)
PRED-TMR PRED-TMR
http://athina.biol.uoa.gr/PRED-TMR/input.html
Pasquier C, Promponas VJ, Palaios GA, Hamodrakas JS, Hamodrakas SJ (1999)
A novel method for predicting transmembrane segments in proteins based on a statistical analysis of the SwissProt database: the PRED-TMR algorithm. Protein Engineering, 12, 381-385. PMID 10360978
(uses statistical analysis of protein database)
proT PRO-TMHMM (in the TOPCONS server)
http://topcons.cbr.su.se/
Viklund H, Elofsson A (2004)
Best alpha-helical transmembrane protein topology predictions are achieved using hidden Markov models and evolutionary information. Protein Sci, 13, 7, 1908-1917. PMID 15215532
Bernsel A, Viklund H, Hennerdal A, Elofsson A (2009)
TOPCONS: consensus prediction of membrane protein topology. Nucleic Acids Research, Web Server Issue 37, W465-W468. PMID 19429891
(uses sequence alignment of membrane protein sequences and hidden Markov model (HMM) trained on membrane protein sequences)
prodivT PRODIV-TMHMM (in the TOPCONS server)
http://topcons.cbr.su.se/
Viklund H, Elofsson A (2004)
Best alpha-helical transmembrane protein topology predictions are achieved using hidden Markov models and evolutionary information. Protein Sci, 13, 7, 1908-1917. PMID 15215532
Bernsel A, Viklund H, Hennerdal A, Elofsson A (2009)
TOPCONS: consensus prediction of membrane protein topology. Nucleic Acids Research, Web Server Issue 37, W465-W468. PMID 19429891
(uses sequence alignment of membrane protein sequences and hidden Markov model (HMM) trained on membrane protein sequences)
SCAMPI SCAMPI
http://scampi.cbr.su.se/
Bernsel A, Viklund H, Falk J, Lindahl E, von Heijne G, Elofsson A (2008)
Prediction of membrane-protein topology from first principles. Proc. Natl. Acad. Sci. USA. 105, 7177-7181. PMID 18477697
(uses biophysical residue hydropathy values and principles of translocon functioning)
SCAMPImaT SCAMPI-multi (multiple sequence alignment) (in the TOPCONS server)
http://topcons.cbr.su.se/
Bernsel A, Viklund H, Falk J, Lindahl E, von Heijne G, Elofsson A (2008)
Prediction of membrane-protein topology from first principles. Proc. Natl. Acad. Sci. USA. 105, 7177-7181. PMID 18477697
Bernsel A, Viklund H, Hennerdal A, Elofsson A (2009)
TOPCONS: consensus prediction of membrane protein topology. Nucleic Acids Research, Web Server Issue 37, W465-W468. PMID 19429891
(uses sequence alignment of membrane protein sequences, biophysical residue hydropathy values and principles of translocon functioning)
SCAMPIsqS SCAMPI-sequence (in the TOPCONS-single server)
http://single.topcons.net
Bernsel A, Viklund H, Falk J, Lindahl E, von Heijne G, Elofsson A (2008)
Prediction of membrane-protein topology from first principles. Proc. Natl. Acad. Sci. USA. 105, 7177-7181. PMID 18477697
Hennerdal A, Elofsson A (2011)
Rapid membrane protein topology prediction. Bioinformatics 27, 9, 1322-1323. PMID 21493661
(uses biophysical residue hydropathy values and principles of translocon functioning)
SCAMPIsqT SCAMPI-sequence (in the TOPCONS server)
http://topcons.cbr.su.se/
Bernsel A, Viklund H, Falk J, Lindahl E, von Heijne G, Elofsson A (2008)
Prediction of membrane-protein topology from first principles. Proc. Natl. Acad. Sci. USA. 105, 7177-7181. PMID 18477697
Hennerdal A, Elofsson A (2011)
Rapid membrane protein topology prediction. Bioinformatics 27, 9, 1322-1323. PMID 21493661
Bernsel A, Viklund H, Hennerdal A, Elofsson A (2009)
TOPCONS: consensus prediction of membrane protein topology. Nucleic Acids Research, Web Server Issue 37, W465-W468. PMID 19429891
(uses biophysical residue hydropathy values and principles of translocon functioning)
SOSUI SOSUI
http://bp.nuap.nagoya-u.ac.jp/sosui/sosuiG/sosuigsubmit.html
Hirokawa T, Boon-Chieng S, Mitaku S (1998)
SOSUI: classification and secondary structure prediction system for membrane proteins. Bioinformatics, 14, 378-379. PMID 9632836
Mitaku S, Hirokawa T (1999)
Physicochemical factors for discriminating between soluble and membrane proteins: hydrophobicity of helical segments and protein length. Protein Engineering, 11, 953-957. PMID 10585500
Mitaku S, Hirokawa T, Tsuji T (2002)
Amphiphilicity index of polar amino acids as an aid in the characterization of amino acid preference at membrane-water interfaces. Bioinformatics, 18, 608-616. PMID 12016058
(uses biophysical residue hydropathy values and other physico-chemical properties)
SPLIT4 SPLIT4
http://split.pmfst.hr/split/4/
Juretic D, Zoranic L, Zucic D (2002)
Basic charge clusters and predictions of membrane protein topology. Journal of Chemical Information and Modeling, 42, 620-632. PMID 12086524
(uses biophysical residue hydropathy values and other physico-chemical properties)
stmhmmS S-TMHMM (in the TOPCONS-single server)
http://single.topcons.net
Viklund H, Elofsson A (2004)
Best alpha-helical transmembrane protein topology predictions are achieved using hidden Markov models and evolutionary information. Protein Sci, 13, 7, 1908-1917. PMID 15215532
(uses hidden Markov model (HMM) trained on membrane protein sequences)
SVMtm SVMtm
http://ccb.imb.uq.edu.au/svmtm/
Yuan Z, Mattick JS, Teasdale RD (2004)
SVMtm: Support vector machines to predict transmembrane segments. Journal of Computational Chemistry, 25, 632-636. PMID 14978706
(uses support vector machine (SVM) trained on membrane protein sequences)
SVMtop SVMtop
http://bio-cluster.iis.sinica.edu.tw/~bioapp/SVMtop/
Lo A, Chiu HS, Sung TY, Lyu PC, Hsu WL (2008)
Enhanced membrane protein topology prediction using a hierarchical classification method and a new scoring function. J Proteome Res. 7, 2, 487-496. PMID 18081245
(uses support vector machine (SVM) trained on membrane protein sequences)
TMAP EMBOSS TMAP
ftp://emboss.open-bio.org/pub/EMBOSS/EMBOSS-6.4.0.tar.gz (TMAP was run without any sequence alignment inputs.)
Persson B, Argos P (1994)
Prediction of transmembrane segments in proteins utilising multiple sequence alignments. Journal of Molecular Biology, 237, 182-192. PMID 8126732
Persson B, Argos P (1996)
Topology prediction of membrane proteins. Protein Science, 5, 363-371. PMID 8745415
(uses sequence alignment of membrane protein sequences)
TMHMM2 TMHMM Server v. 2.0
http://www.cbs.dtu.dk/services/TMHMM/
Sonnhammer ELL, von Heijne G, Krogh A (1998)
A hidden Markov model for predicting transemembrane helices in protein sequences. Proceeding of Sixth International Conference on Intelligent Systems for Molecular Biology,
Vol. 5, AAAI/MIT Press, Menlo Park, CA, pp. 175-182. PMID 9783223
Krogh A, Larsson B, von Heijne G, Sonnhammer ELL (2001)
Predicting transmembrane protein topology with a hidden Markov model: application to complete genomes. Journal of Molecular Biology, 305, 567-580. PMID 11152613
(uses hidden Markov model (HMM) trained on membrane protein sequences)
TMLOOP TMLOOP
http://membraneproteins.swan.ac.uk/TMLOOP
Lasso G, Antoniw JF, Mullins JGL (2006)
A combinatorial pattern discovery approach for the prediction of membrane dipping (re-entrant) loops. Bioinformatics 22, 14, e290-e297. PMID 16873484
(uses combinatorial pattern discovery trained on membrane protein sequences)
TMMOD TMMOD
http://liao.cis.udel.edu/website/servers/TMMOD/
Kahsay RY, Gao G, Liao L (2005)
An improved hidden Markov model for transmembrane protein detection and topology prediction and its applications to complete genomes. Bioinformatics, 21, 9, 1853-1858. PMID 15691854
(uses hidden Markov model (HMM) trained on membrane protein sequences)
TMPRED TMPRED
http://www.ch.embnet.org/software/TMPRED_form.html
Hofmann K, Stoffel W (1993)
TMbase - a database of membrane spanning proteins segments. Biological Chemistry Hoppe-Seyler, 374, 166. PMID
(uses statistical analysis of protein database)
TOPCONS TOPCONS
http://topcons.cbr.su.se/
Bernsel A, Viklund H, Hennerdal A, Elofsson A (2009)
TOPCONS: consensus prediction of membrane protein topology. Nucleic Acids Research, Web Server Issue 37, W465-W468. PMID 19429891
(uses consensus of results of other prediction methods and sequence alignment of membrane protein sequences)
TOPCONSs TOPCONS-single
http://single.topcons.net/
Hennerdal A, Elofsson A (2011)
Rapid membrane protein topology prediction. Bioinformatics 27, 1322-1323. PMID 21493661
Bernsel A, Viklund H, Hennerdal A, Elofsson A (2009)
TOPCONS: consensus prediction of membrane protein topology. Nucleic Acids Research, Web Server Issue 37, W465-W468. PMID 19429891
(uses consensus of results of other prediction methods)
TOPPRED2 TOPPRED2
ftp://ftp.pasteur.fr/pub/gensoft/projects/toppred/toppred-1.10.tar.gz
von Heijne G (1992)
Membrane protein structure prediction, hydrophobicity analysis and the positive-inside rule. Journal of Molecular Biology, 225, 487-494. PMID 1593632
Claros MG, von Heijne G (1994)
TopPred II: an improved software for membrane protein structure predictions. Comput Appl Biosci. 10, 6, 685-686. PMID 7704669
(uses biophysical residue hydropathy values and other physico-chemical properties)
VALPRED VALPRED
http://sydney.edu.au/pharmacy/sbio/software/valpred.shtml algorithm=VALPRED
(publication in preparation)
(uses threading and biophysical residue hydropathy values and solvent accessible surface area)
VALPRED2 VALPRED2
http://sydney.edu.au/pharmacy/sbio/software/valpred.shtml algorithm=VALPRED2
(publication in preparation)
(uses threading and biophysical residue hydropathy values and solvent accessible surface area)
waveTM waveTM
http://bioinformatics.biol.uoa.gr/waveTM
Pashou EE, Litou ZI, Liakopoulos TD, Hamodrakas SJ (2004)
waveTM: wavelet-based transmembrane segment prediction. In Silico Biol, 4, 2, 127-131. PMID 15107018
(uses biophysical residue hydropathy values and dynamic programming algorithm)


SOURCE CODE : benchmark_pl.txt
Archive file of data and server files : benchmark.tgz
(To unarchive the archived file, copy to Linux operating system and type the following : tar -xvf benchmark.tgz )
CONTACT : erat8805 AT uni . sydney . edu . au


COPYRIGHT :
Copyright © 2012 Emma M. Rath. All rights reserved.
Soon this program will be released under an open source software license such as GNU General Public License or
Creative Commons license for Free Software Foundation's GNU General Public License at creativecommons.org