Full text of DOI:10.1093/bioinformatics/16.12.1157

Vol. 16 no. 12 2000
Pages 1157–1158
BIOINFORMATICS APPLICATIONS NOTE
PANAL: an integrated resource for Protein
sequence ANALysis
∗
Kevin A. T. Silverstein , Alan Kilian, John L. Freeman, James E.
∗
Johnson, Ihab A. Awad and Ernest F. Retzel
Computational Biology Centers, Academic Health Center, University of Minnesota,
Mayo Mail Code 43, 420 Delaware St SE Minneapolis, MN 55455-0312, USA
Received on May 31, 2000; revised on August 21, 2000; accepted on August 23, 2000
Abstract
plications: (1) A user must manually input sequences at
Summary: We present PANAL, an integrated resource for
protein sequence analysis. The tool allows the user to
simultaneously search a protein sequence for motifs from
several databases, and to view the result as an intuitive
graphical summary.
Availability: PANAL can be freely accessed on our
World Wide Web server: http://mgd.ahc.umn.edu/panal/.
Source code is available upon request to non-profit
organizations.
each web site; (2) the results from the sites are in a vari-
ety of incompatible formats; and (3) consequently, no in-
tegrated summary of the results from all methods can be
provided.
We have created PANAL to address these issues.
PANAL is an integrated resource that searches for motifs
from PROSITE (both patterns and profiles), PRINTS,
Blocks, and Pfam concurrently. The following public pro-
grams are used, respectively, for the searches: prosite,
pfscan, fingerPRINTScan, blimps/blkprob, and
hmmpfam. In addition to the text reports provided by
each program, we have developed a consistent graphical
summary of the results of all methods.
The main PANAL webpage allows the user to enter a
sequence into a form, and to select motif databases to
search against. The user may filter out matches based
on statistical significance. Once the user submits the
sequence, a Tcl script runs each selected program on
our servers. The only non-default parameters sent to the
programs are the E-value cutoff provided by the user
and the current primary sequence database size used in
the computation of E-values. E-values reported from all
programs are re-normalized, for consistency, to the same
database size (see http://mgd.ahc.umn.edu/panal/e values.
html). Status indicators notify the user which programs are
still in progress. As each program completes, a java applet
dynamically integrates a summary of the results into the
display.
Contact: panal@ahc.umn.edu
Sequence-similarity programs such as BLAST (Altschul
et al., 1990) and FASTA (Pearson and Lipman, 1988)
provide useful insights on the general function of new
protein sequences. However, several methods exist with
higher sensitivity and reliability (Bork and Gibson, 1996;
Eddy, 1996; Hofmann, 1998). Known collectively as
family-based methods, these programs extract features
from a multiple sequence alignment of related sequences,
and use a model of the conserved features to find new
members (Bork and Gibson, 1996; Hofmann, 1998).
Among the models, in increasing complexity, are: (i)
regular-expression patterns, used by PROSITE (Hofmann
et al., 1999); (ii) groups of potentially far-separated
ungapped position specific scoring matrices (PSSMs),
employed by PRINTS (Attwood et al., 2000) and Blocks
(
Henikoff et al., 2000); (iii) profiles, which are essentially
PSSMs with gaps, also used by PROSITE; and (iv) Hidden
Markov Models (HMMs), the statistical model used by
Pfam (Bateman et al., 2000).
Each of these family-based methods has its own merits.
Pfam is well-designed to detect distantly-related family
members. PRINTS is capable of identifying subtle sub-
family relationships. Each method covers a slightly dif-
ferent, but overlapping, set of families. Considering this
complementarity, it is best to scan sequences of interest
using each method in turn. This approach has its own com-
An example of the graphical summary for the sequence
GTPA HUMAN is provided in Figure 1. More motifs
are identified in this sequence than in a typical query
sequence. However, this sequence is useful to illustrate the
differences among the programs.
In the display, a horizontal line representing the user’s
input sequence stretches across the top, along with vertical
lines that indicate amino acid position in the sequence.
Glyphs for each matched motif are shown below the
user’s sequence. These glyphs are color- and shape-coded
∗
Authors to whom correspondence should be addressed.
ꢀ
c Oxford University Press 2000
1157

K.A.T.Silverstein et al.
Fig. 1. A snapshot of PANALs graphical summary report for the sequence GTPA HUMAN. Each motif-matching program is represented by
a different color and shape. Results are sorted by expected value. Additional details are provided in the text.
according to the searching program they represent, and are
sorted vertically by expect value.
References
Altschul,S.F., Gish,W., Miller,W., Myers,E.W. and Lipman,D.J.
(1990) Basic local alignment search tool. J. Mol. Biol., 215, 403–
410.
Databases that match multiple motifs in a family
signature, such as PRINTS and Blocks, are displayed
as a connected set of shapes. A horizontal black bar
indicates the extent of the complete family signature.
Matches to individual motifs in the signature have been
labeled strong/weak for quick visual inspection. The
fingerPRINTScan program (Scordis et al., 1999) defines
strong matches as those with a score greater than 30.
For other programs that generate multiple motifs, we
define strong matches as those whose E-value exceeds
the user-specified E-value cutoff. Strong matches to one
of these motifs are indicated by a rectangle in front of
the connecting bar; weak matches appear as rectangles
behind the bar; missed matches appear as ovals behind
the bar. The shape boundaries indicate the start and end
amino acid positions reported by each matching program.
All glyphs are annotated by a family descriptor and an
expect value, where applicable.
Apweiler,R., Atwood,T.K., Bairoch,A., Birney,A.B.E., Biswas,M.,
Bucher,P., Cerutti,L., Corpet,F., Croning,M., Durbin,R., Fal-
quet,L., Fleischmann,W., Gouzy,J., Hermjakob,H., Hulo,N.,
Jonassen,I., Kahn,D., Kanapin,A., Karavidopoulou,Y., Lopez,R.,
Marx,B., Mulder,N.J., Oinn,T.M., Pagni,M., Servant,F.,
Sigrist,C.J.A. and Zdobnov,E.M. (2000) InterPro—an inte-
grated documentation resource for protein families, domains,
and functional sites. Bioinformatics, 16, in press.
Attwood,T.K.,
Croning,M.D.R,
Flower,D.R.,
Lewis,A.P.,
Mabey,J.E., Scordis,P., Selley,J.N. and Wright,W. (2000)
PRINTS-S: the database formerly known as PRINTS. Nucleic
Acids Res., 28, 225–227.
Bateman,A., Birney,E., Durbin,R., Eddy,S.R., Howe,K.L. and
Sonnhammer,E.L.L. (2000) The Pfam protein families database.
Nucleic Acids Res., 28, 263–266.
Bork,P. and Gibson,T.J. (1996) Applying motif and profile searches.
Meth. Enzymol., 266, 162–184.
Eddy,S.R. (1996) Hidden markov models. Curr. Opin. Struct. Biol.,
Mirror software is used to ensure that all databases and
programs are kept up-to-date. Details on all programs,
databases and procedures are provided at http://mgd.ahc.
umn.edu/panal/databases. A subset of the features and
databases in PANAL may also be found on the InterPro
server (Apweiler et al., 2000).
6
, 361–365.
Henikoff,J.G., Greene,E.A., Pietrokovski,S. and Henikoff,S. (2000)
Increased coverage of protein families with the Blocks database
servers. Nucleic Acids Res., 28, 228–230.
Hofmann,K. (1998) Protein classification and functional assign-
ment. Trends Guide to Bioinformatics, 18–21.
Hofmann,K., Bucher,P., Falquet,L. and Bairoch,A. (1999) The
PROSITE database, its status in 1999. Nucleic Acids Res., 27,
Acknowledgements
This work was supported by an SCA with the USDA-ARS,
grants from NSF (9872565, 9872633), DOE (DE-FC07
2
15–219.
Pearson,W.R. and Lipman,D.J. (1988) Improved tools for biological
sequence comparison. Proc. Natl. Acad. Sci. USA, 85, 2444–
9
7ID13550), USDA (58-3625-8-117), and the University
2
448.
of Minnesota Academic Health Center. Some hardware
used in this work was supplied by an AEG from Sun
Microsystems Inc.
Scordis,P., Flower,D.R. and Attwood,T.K. (1999) Finger-
PRINTScan: intelligent searching of PRINTS motif database.
Bioinformatics, 15, 799–806.
1
158