Version 1.3.13 Release 7/5/2020 *Fixed bug in handling situation when last time point in one replicate is missing *Prints to console message when running Gene Ontology enrichment instead of usual STEM cluster since only one or two tab delimited columns *Fixed command line usage message in STEM Version 1.3.12 Release 6/23/2019 *Updated STEM to handle changes to the GO which had caused some download links to be broken *Slight adjustments to GUI to remove some parts being cutoff in recent Java versions *Clarified in manual batch mode still needs to be run with a computer with a display Version 1.3.11 Release 12/26/2016 *Fixed issue in which profile interface was appearing as blank without resizing the window on macs Version 1.3.10 Release 11/24/2016 *Fixed bug when running a batch GO analysis ('-o option included') that could cause no output to be produced in certain cases. Version 1.3.9 Release 9/10/2016 *Changed license on STEM so now it is GPL v3.0 *updated/added/removed entries and filenames in the GO list to be current *removed automatic download of gene locations for chromosome viewer since was not supported by the site previously downloading it from *fixed a bug causing the 'Name' field in the chromosome viewer not to be parsed. Only 'ID' and 'Alias' fields were parsed *fixed an obsucre bug that occurs when sorting profiles when doing GO analysis and sorting profiles based on few genes Version 1.3.8 Release 2/24/2012 *Added the ability to save the main interface as a svg by clicking the disk icon and giving a .svg extension *Removed default cross-references as no longer valid Version 1.3.7 Release 1/1/2011 *Added fold enrichments to GO tables. This is the observed number of genes divided by the expected. *Added a feature to run GO enrichments for a batch set of genes and output to a text file. This can be done with the flag -o GOoutfile.txt when launching STEM from a command line. All the gene sets should be in one file with a '#' header line to denote breaks between gene sets. GOoutfile.txt is the file where the GO output will be written. *Updated list of GO annotations available for download *Filter NOT modified annotations in annotation files (effects a very small percentage of annotations) *Behind the scenes reorganization of code Version 1.3.6 Release 7/31/2009 * Changed annotation downloads from the geneontology from being ftp based to being http to fix connection problems. Version 1.3.5 Release 7/11/2009 Updated the GO and Cross-reference list. Human cross reference file is no longer available for automatic download. Also specifying location files based on GO category selections is no longer the default action of the interface. All changes were in stem\ST.java Version 1.3.4 Release 1/15/2009 Source code released. Behind the scenes reorganization of code. Version 1.3.3. Release 8/10/2008 *Fixed bug that caused profile vector not to be saved correctly in the headerwhen saving a gene table. *Fixed bug causing gene chromosome locations not to be downloaded correctly for some species. *Removed ipi and vega from biomart_species.txt file. Version 1.3.2 Release 5/5/2008 *Fixed bug that sometimes caused repeat files not to be listed on repeat interface when loading them from a defaults file through STEM interface Version 1.3.1 Release 3/31/2008 *Updated STEM so that it could parse the latest gene ontology files which now have the regulates relationship between some categories. Version 1.3 Release 2/18/2008 1. Added a chromosome viewer integrated into STEM - see manual for more details 2. Added the option to save settings and load them directly into STEM. Settings are saved through the disk icon on the STEM interface, and then loaded into STEM through the "Load Saved Settings" button on the input interface. 3. Added the option to display y-axis tick marks on the main input interface when plotting genes. Also added the option not to display the profile ID, model profile, line, or ordering details on the main interface. 4. Added the option to run STEM in batch from the command line 5. Corrected the STEM documentation on the filtering "Minimum Correlation between Repeats" parameter, in the case of three or more repeats the criteria is based on the pairwise mean, not the median 6. Fixed bug that prevented real time spacing from being recogonized in the defaults file 6. Removed the "-ms512M" option to start STEM from the cmd files Version 1.2.2b Released 10/13/2006 1. Fixed bug in the display of p-values that were extremely close to 0 (<~10^-300), but not exactly 0. In Java 1.5 they were being displayed as Infinity to some large negative power, and in Java 1.4 they were being displayed as a really long number to some power. Version 1.2.1b Released 9/18/2006 1. A memory leak was fixed which previously could have caused STEM to run slower when executing it multiple times without restarting Java Version 1.2.0b Released 9/5/2006 Added several interface options which can be accessed on the main interface window (see also http://www.cs.cmu.edu/~jernst/stem/newoptions.html). These options allow: 1. All non-filtered genes expression profiles to be displayed on the main interface 2. Just gene expression profiles of a given GO category or gene set to be displayed on the main interface 3. The x-axis on the main interface and detail profile windows (cluster windows for k-means) to be scaled based on the real sampling rate, previously all time points could only be uniformly spaced 4. The scale of y-axis of the detail profile windows (cluster windows for k-means) to be fixed to user defined parameters, previously the only option would be for STEM to determine each one automatically Version 1.1.3b Released 8/13/2006 1. Added buttons to all tables to copy the content of the tables to the clipboard. Also can copy just the gene names to the clipboard for gene tables. 2. Fixed a bug in the printing of the header of GO Category gene tables when saving to a file 3. Added back to the list of GO annotations Glossina morsitans (Sanger GeneDB). Added Pseudomonas aeruginosa PA01 (PseduoCap) to the list of GO annotation sources. Version 1.1.2b Released 4/26/2006 1. Minor bug fixed in computing expected counts. If a gene could match multiple profiles equally well during a permutation, which usually only happens if genes with missing data are not filtered, then the expected counts were slightly lower than they should have been. Version 1.1.1b Released 4/25/2006 1. STEM now has a new default gene filtering criteria. Under the new filtering criteria called 'Maximum-Minimum' a gene is filtered if the difference between its largest value and its smallest value after transformation (Log normalize data, Normalize data, or No normalization/add 0) is less than the value of the 'Minimum Absolute Expression Change Parameter'. The original filtering criteria, based on the absolute expression change from time point 0, is now called 'Difference from 0' and can be specified under the 'Change should be based on' option. The 'Maximum-Minimum' filtering criteria has the property that the set of genes that would be filtered is invariant to permuting the order of time points, which is not necessarily true under the 'Difference from 0' filtering criteria. The 'Maximum-Minimum' filtering criteria is thus arguably a more appropriate filtering criteria to use in conjunction with STEM's permutation test. 2. STEM now has the option under Advanced Options-->Model Profiles for its permutation test not to permute time point 0. The permutation test is used to computed the expected number of genes assigned to a profile. Previously and still the default option in STEM is for the permutation test to compute the expected counts based on a permutation of all time points. This is generally the preferred method since it can detect important changes between time point 0 and the immediate next sampled time point. However for some experimental designs particularly two channel experiments where time point 0 serves as the reference sample, not permuting time point 0 is also useful. The reason is that the asymmetry in the experimental design will allow a gene's time point 0 value expression value before any transformation to be known with lower variance than the other time points, and this difference can cause some profiles to be more likely to be significant when permuting time point 0. When not permuting time point 0, profiles found significant are significant independent of the asymmetry in the experimental design between time point 0 and the other time points. 3. A bug was fixed that caused some genes not to be filtered that should have been based on the absolute expression change parameter. This bug only occurred when allowing genes with missing data, choosing log normalize or normalize data, and having no repeats. Only genes with missing data could have incorrectly slipped through filter, and only a smaller percent would probably actually have done so. 4. The error messages for parsing the defaults file was improved in a few places. Version 1.1.0b Released 3/20/2006 1. STEM now provides an implementation of k-means algorithm in addition to the still default STEM clustering method designed for short time series gene expression data published in the Bioinformatics/ISMB 2005 paper. 2. Extremely small p-values are computed almost exactly, previously any p-value below 10^-12 was returned as 0. 3. Added Cow (EBI) to the list of GO annotation sources and removed Glossina morsitans (Sanger GeneDB) which was no longer active on the GO ftp site 4. When a table window is maximized all extra window area is now given to the table instead of some to the button panel 5. When comparing two data sets it is now possible to automatically load in the define gene set dialog box the set of genes filtered in the other data set by setting the profile ID value to "-1" and pressing "Select Genes". 6. Changed the default start memory option for STEM from -ms1024M to -ms512M in the STEM command files. Version 1.0.7b Released 1/5/2006 1. When comparing data sets there is again the option to display gene and GO tables based on the set of genes in the intersection of two profiles. When querying a user defined gene set there is again the option to display gene and GO tables based on the set of genes in the intersection of the user defined gene set and the set of genes assigned to a profile or cluster of profiles. These options are described in the user manual, and appeared in STEM before the release of version 1.04b when they were inadvertently removed. 2. If the window containing the graph of the genes assigned to a model profile is resized, then the graph of the genes is now rescaled accordingly. 3. If the "Order Profiles By" window or the "Order Clusters By" window is open, but minimized, pressing its corresponding button on the main interface now makes the window visible again. Previously there would be no response when pressing the button in that situation. 4. If STEM cannot save the contents of a table to a file because another application has a write lock on the file, then a new window containing an error message now pops up. Previously the error message was just printed to the console window. Version 1.0.6b Released 11/21/2005 1. Bug fixed that prevented GO results from being displayed on computers that default to a language for which numbers have a ',' in places in the English system where there is a '.' (e.g. 3,21 instead of 3.21). 2. All numbers in tables are now displayed under the English system Version 1.0.5b Released 11/1/2005 1. Bug fixed that could in certain circumstances cause STEM not to use all available Gene Ontology (GO) annotations. If a gene symbol in an official 15 column annotation files had a lower case letter in it and matched a gene symbol in the data file, STEM incorrectly did not use the annotation. This bug however does not effect the EBI cross references. This bug also does not affect users providing their own two column gene annotation files. 2. Removed reference to the defunct GenBank (Compugen), Gene Index (TIGR), and UniProt (Compugen) annotation sources. Update reference to TIGR and TAIR annotations for Arabidopsis thaliana to reflect they are now merged into one file. Version 1.0.4b Released 10/25/2005 1. Bug fixed that caused the advanced options help info windows not to be closeable until the advanced options dialog window was closed. This bug only appears when using Java 1.5, and was not present with Java 1.4. 2. Slight changes to the color scheme for significant model profiles 3. Changed the sample comparison data file included with STEM Version 1.0.3b Released 7/17/2005 1. The parsing algorithm of the gene ontology file was modified internally. The algorithm previously relied on assumptions about the format of the file which is and has been true, but are not documented to necessarily be the case in the future. These changes currently have no visible effect. Also more informative error messages were added in the event there are problems parsing the gene ontology file. Version 1.0.2b Released 7/15/2005 1. Bug fixed in parsing the very last category term in the gene ontology file. This was causing a null pointer exception to be thrown with the settings specified in defaultsGuilleminSample.txt and the latest ontology, but was not noticeable with previous versions of the ontology. Version 1.0.1b Released 7/1/2005 1. Bug fixed allowing the gene table for a GO category to now be saved when doing a GO enrichment analysis on non-time series data. Version 1.0.0b Released 6/24/2005