metaP
degprune
hemeBIND
degPareto

HemeBIND Manual

go back to server


Heme binding residue prediction from protein sequence and structure


///////////////////////////////////////////////////////////////////////////////////////////////////

Methodology

HemeBIND is a novel algorithm for prediction of heme binding residues by combining structure and
sequence information. Given the structure of a protein, HemeBIND extracts several structural and
sequence features of the target residue and its spatial neighbors, including position specific
scoring matrix, relative accessible surface area, depth index and protrusion index. The support
vector machine is used to construct the structure-based classifier and a post-processing procedure
is proposed to reduce the false positives produced by this classifier. For a protein sequence,
HemeBIND provides a sequence-based classifier by using the PSSM of a sliding window centered on
the target residue as the input of SVM. To further improve the prediction, HemeBIND utilizes a voting
method to combine the outputs of the structure-based and sequence-based classifiers.

///////////////////////////////////////////////////////////////////////////////////////////////////

Input Format

For a sequence prediction:

The input of the sequence prediction is an amino acid sequence in Fasta format.

>1O1M:A
MLSPADKTNVKAAWGKVGAHAGEYGAEAFERMFLSFPTTKTYFPHFDLSHGSAQVKGQGKKVADALTNAVAHVDDMPNAL
SALSDLHAHKLRVDPVNFKLLSHCLLVTLAAHLPAEFTPAVHASLDKFLASVSTVLTSKYRGGGVLSPADKTNVKAAWGK
VGAHAGEYGAEAFERMFLSFPTTKTYFPHFDLSHGSAQVKGQGKKVADALTNAVAHVDDMPNALSALSDLHAHKLRVDPV
NFKLLSHCLLVTLAAHLPAEFTPAVHASLDKFLASVSTVLTSKYR

For a structure prediction:

The input of the structure prediction is an X-ray crystal structure in PDB format.

ATOM 1 N MET A 1 19.555 29.903 42.490 1.00 50.76 N
ATOM 2 CA MET A 1 20.614 30.851 42.171 1.00 54.92 C
ATOM 3 C MET A 1 21.995 30.221 42.321 1.00 56.63 C
ATOM 4 O MET A 1 22.237 29.387 43.195 1.00 55.52 O
ATOM 5 CB MET A 1 20.491 32.089 43.048 1.00 60.12 C
ATOM 6 N LEU A 2 22.910 30.635 41.449 1.00 47.67 N
ATOM 7 CA LEU A 2 24.304 30.230 41.561 1.00 39.94 C
ATOM 8 C LEU A 2 25.108 31.273 42.334 1.00 37.57 C
ATOM 9 O LEU A 2 25.342 32.368 41.821 1.00 58.96 O
ATOM 10 CB LEU A 2 24.921 30.038 40.176 1.00 40.96 C
ATOM 11 CG LEU A 2 24.451 28.823 39.376 1.00 40.06 C
ATOM 12 CD1 LEU A 2 24.794 28.991 37.903 1.00 38.65 C
ATOM 13 CD2 LEU A 2 25.054 27.548 39.941 1.00 31.43 C
ATOM 14 N SER A 3 25.535 30.937 43.543 1.00 31.50 N
ATOM 15 CA SER A 3 26.465 31.776 44.291 1.00 30.86 C
ATOM 16 C SER A 3 27.633 32.236 43.430 1.00 37.53 C
ATOM 17 O SER A 3 27.927 31.657 42.381 1.00 34.29 O
ATOM 18 CB SER A 3 27.032 30.985 45.481 1.00 33.36 C
ATOM 19 OG SER A 3 28.180 30.254 45.070 1.00 29.16 O
ATOM 20 N PRO A 4 28.336 33.276 43.866 1.00 36.14 N
ATOM 21 CA PRO A 4 29.566 33.673 43.178 1.00 30.05 C
ATOM 22 C PRO A 4 30.617 32.573 43.285 1.00 30.97 C
ATOM 23 O PRO A 4 31.385 32.323 42.357 1.00 39.50 O
ATOM 24 CB PRO A 4 30.022 34.918 43.938 1.00 39.49 C
ATOM 25 CG PRO A 4 28.817 35.391 44.679 1.00 43.04 C
ATOM 26 CD PRO A 4 28.015 34.158 44.998 1.00 39.18 C

* Note that the structural file should only cotain the lines that begin with 'ATOM' and
provide at least 19 residue coordinates.

//////////////////////////////////////////////////////////////////////////////////////////////////

Output Format

For a sequence prediction: For a structure prediction:

SeqNo AA Tag StrPos AA Tag
1 M - 1 M -
2 L - 2 L -
3 S - 3 S -
4 P - 4 P -
5 A - 5 A -
6 D - 6 D -
7 K - 7 K -
8 T - 8 T -
9 N - 9 N -
10 V - 10 V -
11 K - 11 K -
12 A - 12 A -
13 A - 13 A -
14 W + 14 W +
15 G - 15 G -
16 K - 16 K -
17 V + 17 V +
18 G - 18 G -
19 A - 19 A -
20 H - 20 H -
21 A - 21 A -
22 G - 22 G -
23 E - 23 E -
24 Y + 24 Y +
25 G - 25 G -
26 A - 26 A -
27 E - 27 E -
28 A + 28 A -
29 F - 29 F -
30 E - 30 E -

* SeqNo: Sequencial number; StrPos: Structural position; AA: Amino acid.
/////////////////////////////////////////////////////////////////////////////////////////////////

Download dataset

////////////////////////////////////////////////////////////////////////////////////////////////

Citing & Credits

When citing the HemeBIND server please refer to:

Rong Liu, Jianjun Hu. (2010) HemeBIND: a novel method for heme binding residue prediction
by combining structural and sequence information. Submitted.

Acknowledgements

Development of HemeBIND was supported by the National Science Foundation (Grant BIO-DBI-0845381).

/////////////////////////////////////////////////////////////////////////////////////////////////

Contact

Dr. Rong Liu
Dr. Jianjun Hu
Department of Computer Science and Engineering
University of South Carolina
Columbia, SC 29208
E-mail: liurong116@tom.com or jianjunh@cse.sc.edu