PROMPT version overview
Current version 0.9.7
- Serialization bug of the generic Sequence-Property class fixed.
- Automatic All-vs-All
Along with the script recording it is possible to add a simple loop construction around the analysis action. Thus you will be able to process large-scale all-versus-all actions and can filter the results for significance as you like.
- Performance Speed-ups
New implementations using advanced caching technologies will allow the usage of large data sets.
Sign up to the community mailinglist to keep informed.
- Swiss/UniProt retrieval updated. Secondary accessions are now replaced by current citeable primary accession code. Additionally the user is warned if an ambiguous accession is found.
- The latest JVM "JRE1.6.0_05" is now compatible with PROMPT. This specific Java version returns -against the specifications- the version string "10.0-b19" in the system-property "java.vm.version".
- On very rare occasions a PROMPT Blast run could hang, this is fixed now.
- File Conversion Tools added
- Ignore-All-Errors variable is now public and can be set in your automated pipelines. (Example Bean-Shell code)
- Complete re-design of the Swiss-Prot comparison engine. The Swiss-prot/UniProt version 2 input classes can now used with all generic comparison methods.
- Added support for SIMAP sequence identifiers.
- Improved usability: tab-delimited text import allows ignore-all or cancel on errors now, Generic-Subset Enrichment Analysis checks if the subset is drawn from the population in advance, Errors that are caused by remote Web Services are communicated more clearly to the user.
- Scripting mode improved: importing tab-delimited text files using PROMPT Beanshell scripts now uses utf-8 encoding instead using the underlying system encoding. This should increase plattform-independence
- Binning information stored as additional results. Now it is possible to see which genes/proteins are put into which bins in a numerical or categorical comparison
- PEDANT access updated and performance improvements
- Some serialization problems are fixed now.
- New: Macro-Script
recording and executing
All actions that are performed interactively in the GUI can be recorded. The resulting scripts are BeanShell scripts and can be run interactively as well as in a batched-fashion. Furthermore you can access all Java classes and method as well as the complete functionality of the PROMPT framework from the scripts. The scripts do not need any compilation and can be executed immediately.
Minor incompatibility with the new BLAST version 2.2.15 on some LINUX boxes fixed.
Extension of the PROMPT Fast Mapping Method: Along of each hit you'll get information about the taxonomy and the database where the hit was found. This is especially useful if you "blast" against all PEDANT databases or against a metadatabase like PDB or NCBI COG.
Fixed minor bug in the PROMPT Fast Mapping metho
- Improved run time of the Mann-Whitney test
- New: Access and integration
of all PEDANT resources.
You will be able to retrieve all annotations available in PEDANT for your proteins of interest and compare, export, merge or integrate it with your external data. All PROMPT methods, visualisations and statistics can be used along with the PEDANT data. Screen shots of the PROMPT-PEDANT integration dialog.
- All parameter that are used in a BLAST run can be changed now. Additionally error messages that may occur during a BLAST run will be prompted to the user in an interactive dialog where it is possible to ignore once, always or cancel the job. (A change of the allowed parameters in the new BLAST version 2.2.14 made this step necessary)
- New: PROMPT Fast-Mapping (PROMPT-FM)
The PROMPT Fast--Mapping Method allows you to "blast" any protein sequences to really large or gigantic databases up to 50 times° faster than a single conventional BLAST run. All standard databases like GenBank, UniProt, RefSeq, PDB and all (>400) PEDANT databases can be searched. No download of any of these databases is necessary. Another advantage is that no local BLAST installation is necessary. Performance Evaluation
New: RegEx - Easy-menu driven application of regular expressions on/within the result data. More about this features.
- New: R-independent mode: PROMPT can start now even if no R/S software is installed. The frequently used tests like Mann-Whitney (MW), Kolmogorov-Smirnov, Pearson Correlation & Test and the Chi-Square Test are added as JAVA implementation. However, the interactive test dialogs and the R/S visualisations will be disabled.
(° The Fast-Mapping performance evaluation depends on your CPU power, the internet connection and the database size that should be searched. The speed-up is only significant in searches against really large datasets)
- Minor bugs fixed,
- BioJava and other exception catched more gracefully,
- Improved memory usage: all-in-memory XML-DOM-tree are replaced with low-memory-footprint XML-Stream technologies
- Inital public release