I-TASSER SUITE (Copyright 2015 by Zhang Lab, University of Michigan, All rights reserved) (Version 4.4, released on 2015/09/02) ------------------------------------------------------- What is new in I-TASSER V4.4 compared to I-TASSER V4.3? ------------------------------------------------------- 1. Fixed a few bugs in 'download_lib.pl' for failure of library download in some system. 2. Corrected errors in directory setting in 'mMKRESmod' 3. Fixed bugs in file copy and parameter setting in 'rerankINIT.pl', and 'zysubmod'. 4. A new Monte Carlo simulation program was added (casres) for using user specified restraint information. 5. README.txt was updated with a few changes on command line for running I-TASSER. ------------------------------------------------------- What is new in I-TASSER V4.3 compared to I-TASSER V4.2? ------------------------------------------------------- 1. A new version of ResQ program was integrated for estimating B-factor and the residue-level errors of protein structure prediction. 2. Fixed a few bugs in the I-TASSER Monte Carlo simulation program to tolerate unphysical template alignments. 3. Fixed a bug in 'checkinput.pl' to check solvation prediction errors. 4. Several new scripts were updated to facilate I-TASSER implementation. ------------------------------------------------------- What is new in I-TASSER V4.2 compared to I-TASSER V4.1? ------------------------------------------------------- 1. Source codes of following programs were released in the 'src' folder: (a) wdPPAS (b) wMUSTER (c) wPPAS (d) pair99 (e) EDTSurf (f) COACH (g) TM-SITE (h) S-SITE (i) COTACTOR (j) TM-align_search2 (k) PRTMalign2 (l) EClocal (m) HAAD (n) REMO (o) addchinid (p) ca2sg (q) addmysg2 (r) SPICKER (s) TM-align (t) TM-score (u) NW-align 2. Some bugs were fixed as reported by users of I-TASSER V4.1. (a) Fix a bug in 'get_rsq_bfp.pl' when number of templates is <100 (b) 'COFACTORmod' updated to deal with missing templates for 'Lsearch2' (c) A bug in 'runCOFACTOR.pl' and 'runTMSITE.pl' was fixed. (get_absolute_path) (d) A bug in 'COACHmod' was fixed. The COFACTOR output path was overwritten before copying the results to datadir. (e) A bug in 'COCOFACTORmod' was fixed. Correct cases when lig_* not available: `cp lig_*.pdb $BSrecorddir/`. (f) Instructions for using other job management software (rather than the default PBS) are provided at: http://zhanglab.ccmb.med.umich.edu/bbs/?q=node/3561 ------------------------------------------------------- What is new in I-TASSER V4.1 compared to I-TASSER V4.0? ------------------------------------------------------- 1. A new package ('update_IT_lib.tar.bz2') was provided which allows users to create template files on their own from any protein structures. 2. The script 'download_lib.pl' was upgraded with following changes: (a) A new option (-N) was added to allow users to decide if they want to download the non-redundant sequence database; (b) Allow user to estimate the template library size before download. 3. Additional options were added to the 'runI-TASSER.pl' and other scripts which allow users to specify I/O paths to any directories, including -datadir: directory of the input files -pkgdir: directory of the I-TASSER package -libdir: directory of the template library -outdir: directory of the output files (default value is datadir) Both absolute and relative paths (e.g. '/home/username' or '../username') are acceptable now for these paths. ------------------------------------------------------- What is new in I-TASSER V4.0 compared to I-TASSER V3.0? ------------------------------------------------------- 1. Estimation of residue-specific accuracy in I-TASSER models was added. 2. Prediction of normalized B-factor was added. 3. A script 'show_align.sh' for visualizing the threading alignments was added. (Adapted from Dr. Stephane M. Gagne) 4. Some bugs in 'runI-TASSER.pl' were fixed as reported by users of I-TASSER V3.0. (a) The option '-walltime' was replaced by '-hours' to allow users to specify maximum running time in the simulation. This is to avoid confusion with the walltime option in qsub command. (b) Running of the HAAD program was skipped as many users' clusters do not have libgfortran.so installed. 5. A few bugs in the I-TASSER simulation core program was fixed, including the one that cannot run with screwed initial whole-chain model. ------------------------------------------------------- What is new in I-TASSER V3.0 compared to I-TASSER V2.1? ------------------------------------------------------- 1. Four new protein function prediction programs were added: (1) COACH for ligand-binding site, EC number and GO terms prediction. (2) COFACTOR for ligand-binding site, EC number and GO terms prediction. (3) TM-SITE for ligand-binding site prediction. (4) S-SITE for ligand-binding site prediction. 2. Template libraries and programs were separated into independent locations for users' convenient update of the package. 3. A script 'download_lib.pl' was provided for easy download of all library files. 4. The names of threading programs were standardized to: PPAS, dPPAS, dPPAS2, Env-PPAS, MUSTER, wPPAS, wdPPAS, wMUSTER. 5. An option of '-walltime' was added in 'runI-TASSER.pl' to allow users to specify walltime in the simulation. Three new options for function prediction ('LBS', 'EC' and 'GO' were added as well. 6. Some bugs in 'runI-TASSER.pl' were fixed as reported by users of I-TASSER V2.1. ------------------------------------------------------- What is new in I-TASSER V2.1 compared to I-TASSER V2.0? ------------------------------------------------------- 1. An option was added to allow users to specify JAVA path in their machine. 2. A number of bugs were fixed including (a) missed program 'addchainid' was added in; (b) threading names are unified in 'get_cscore.pl'. ------------------------------------------------------- What is new in I-TASSER V2.0 compared to I-TASSER V1.0? ------------------------------------------------------- 1. Protein secondary structure prediction is improved. In I-TASSER V1.0 package, PSI-pred was used to predict secondary structure of target proteins. PSI-pred is third-party software and users need to download the program from other website, which often creates incompatiobility and mistakes in I-TASSER suite configurations. We recently developed a novel secondary structure prediction (PSSpred) by combining seven neural network predictors from different profile data and parameters. The Q3 accuracy of PSSpred is ~3% higher than that of PSI-pred. Moreover, PSSpred is seamlessly integrated with the I-TASSER Suite configurations. 2. New threading programs are integrated. LOMETS in I-TASSER V1.0 includes five threading programs for protein fold- recognition and template identification. In I-TASSER V2.0, three new threading programs with a novel sequence profile generation stratagy are added. This helps I-TASSER to detect more broad range of structure templates for the I-TASSER fragment assembly. Accordingly, the average TM-score of the final models in I-TASSER V2.0 is ~5% higher than that by I-TASSER V1.0 based on the same template library. 3. More flexible for user-intervened modeling. Users sometimes have the background or structural information of their targets from experimental or biological analysis. Such information is very valuable for protein structure prediction. In I-TASSER V2.0, users are allowed to specify template, target-template aligments, residue- residue contact/distance information to guide the I-TASSER simulations. I-TASSER V2.0 also allows users to exclude specific template structures and/or the templates homologous to the target at certain distance cutoff. This is often useful for different purpose, e.g. for algorithm benchmark and comparisons. These features are newly added to I-TASSER V2.0. 4. A "light" option is added to allow faster modeling. A new option "-light" was implemented in I-TASSER V2.0, which sets the maximum time of folding simulation below 5 hours. It can be used for the users who wants quick results for large easy targets. For large hard targets, full version simulations are needed for the best results. 5. Provide an option for trajectory file deposit. Advanced users may be interesed in I-TASSER trajectory data analysis. A new option "-traj" was added in I-TASSER V2.0 to allow users to deposit structural decoys from I-TASSER simulations. 6. Fixing bugs in installation and simulations. A number of bugs in the installation and implementation, as reported by uses of I-TASSER V1.0, have been fixed in I-TASSER V2.0. Please post your questions and comments at http://zhanglab.ccmb.med.umich.edu/bbs/?q=forum/2