We've created the following software packages. To obtain most of the scientific packages, you must fill out a license. Most programs are freely available to academic institutions, but we would still like you to complete a license so that we know who is using our programs. The easiest way to get a license for an academic site is to go to Academic Software Licensing. If this fails to do what you want, or you would like to discuss a commercial license, or you need a license for something not included there, please contact us about other licensing agreements.
For more information on our current research and software projects, please see our research pages.
The backbone-dependent rotamer library and a program for its use provide a fast means of predicting sidechain conformation from backbone phi and psi values. These methods can be used for homology modeling and structure refinement by x-ray and NMR.
The program DRAWBRIDGE builds loop regions onto protein structures. DRAWBRIDGE supercedes an earlier program with similar functionality, named BLOOP. Unlike database or build-up methods, Bloop uses the conformational propensities of amino acids with a genetic algorithm to generate novel candidates for protein loop regions. End constraints are satisfied using a random tweak method.
Availability: DRAWBRIDGE is available under license from the Regents of the University of California.
References: "Conformational Sampling of Loop Structures Using Genetic Algorithms" Christine S. Ring and Fred E. Cohen, Israel Journal of Chemistry (1994) 34(2):xx.
The FOLD suite of programs constructs three-dimensional models of protein structure from known segments of secondary structure. The observed geometries of secondary structure interactions are used to combinatorically generate large numbers of structures that sample the majority of conformational space. Structures are eliminated that fail to meet simple criteria pertaining to the connectivity of the chain or the steric properties of the molecule. Many different protein topologies can be constructed.
Computer Requirements: The original versions of the FOLD suite are written in
FORTRAN77 and should compile on most computers. An updated version that builds
all-alpha helix models is written in C++.
Availability: The Fold suite of programs is available under license from the
Regents of the University of California.
Academic Price: No Charge
Industry Price: $6000
References: "Tertiary Structure Prediction" F. E. Cohen & I. D. Kuntz in
Prediction of Protein Structure and the Principles ofProtein Conformation, G.
D. Fasman, Ed. (Plenum, 1989).
Models of proteins created by
the FOLD programs are available.
JColorGrid is a Java graphics program for transforming numerical or text data into color-grids. A color-grid is a 2-dimensional data array that displays information with an associated color hue. Color-grid representations provide an excellent means of illustrating complex data sets and can be used to identify trends, clusters and extremes in a given data set.
Price: Free to Academic users. Industry: Contact Adelle Lohse for pricing.
JEvTrace is a JAVA implementation of a refined Evolutionary Trace method and its variations. The software is composed of a GUI and underlying algorithms. JEvTrace integrates multiple sequence alignment and phylogenetic data within a graphical and algorithmic framework. Optionally, the software relies on WebMol for visualization of results with respect to protein structures and generation of 3D structure graphics. For details, please refer to the publication:
Joachimiak, M.P., Cohen, F.E.
"JEvTrace: refinement and variations of the evolutionary trace in JAVA."
Price: Free to Academic users. Industry: Contact Adelle Lohse for pricing.
Match is a program which applies the PLANS regular expression pattern matching language to protein sequences. The concept of the user interface is that of a workspace containing a set of sequences and a set of patterns. Files laden with sequences are loaded into the workspace, as are filesof pattern descriptions. The user may view the individual sequences or patterns from within the environment. The pattern matching algorithm canbe applied to the intersection formed by any sequence subset and anypattern subset.
Computer Requirements: Match is written in C and is available for UNIX or VMS
computers. A Macintosh version, MacMatch, is also available from QCPE.
Availability: Match is available under license fromthe Regents of the University
of California.
Academic Price: No Charge
Industry Price: $3000
References: Cohen, F. E., R. M. Abarbanel, I. D. Kuntz, and R. Fletterick, "Secondary
Structure Assignmen for Alpha/Beta Proteins by a Combinatorial Approach," Biochemistry
(October 1983).
Cohen, F. E., R. M. Abarbanel, I. D. Kuntz, and R. Fletterick, "Turn Prediction
in Proteins Using a Pattern Matching Approach", Biochemistry (January 1986).
MINAREA is a program for computing the optimal structural superposition between two protein segments.
nnpredict is a program that predicts the secondary structure type for each residue in an input amino acid sequence. The basis of the prediction is a two-layer, feed-forward neural network. nnpredict can take into account the tertiary class of the protein (either none, all- alpha, all-beta, or alpha/beta) when predicting secondary structure. You can submit your sequence to either the web version.
Pred2ary is a Java program which predicts the secondary structure and structural class of proteins. It runs as an applet or as a standalone application program.
QPack performs a "quick" evaluation of steric packing in protein structures. Amino acid side chains are not represented in full atomic detail. Instead, one or more averaged centers of interaction per residue are used. A measure of the quality of side chain packing is determined from the available spherical space available to each residue. Additionally, an empirical pairwise residue contact potential can be used to estimate the correctness of model protein structures.
References: "A Novel Method for the Rapid Evaluation of Packing in Protein Structures" Lydia M. Gregoret and Fred E. Cohen, Journal of Molecular Biology (1990) 211:959-974.
SCWRL is a program for adding sidechains to a protein backbone based on a backbone-dependent rotamer library. The library provides lists of chi1-chi2 pairs for residues at given phi-psi values, and explores these pairs to try to minimize sidechain-backbone clashes and sidechain-sidechain clashes. Preliminary studies show prediction of chi1 correct within 40 degrees for all residues is 77.0%. Chi3 and chi4 are built in extended conformations.
Availability: SCWRL 2.95 was the last version available under license from the
Regents of the University of California and is no longer supported. The source
code is available for Academic use at no charge.
Contact The Dunbrack Group
for information about SCWRL 3.0.
References: "Prediction of Protein Side-chain Rotamers from a Backbone-Dependent
Rotamer Library: A New Homology Modeling Tool" Michael J. Bower, Fred E. Cohen,
and Roland L. Dunbrack, Jr., Journal of Molecular Biology (1997) 267:1268-1282.
WebMol is a Java program designed to display and analyze structural information contained in the Brookhaven Protein Data Bank (PDB). It runs as an applet or as a standalone application program. It serves demonstration needs in web documents and allows sophisticated structural analysis with its many unique analysis modules; for example, interactive Ramachandran and distance matrix maps.
Scott also has a Homology Modeling with Midas document for UCSF modelers.
(an overview of publicly available systems)
Patches to a number of programs have been developed by Scott Presnell. They are available on the ftp site.
webmaster@cmpharm.ucsf.edu