Today we release Mayday 2.12.

During the development of version 2.12, 100 core builds (#1161-#1260) were produced since December 17, 2010 (102 days). Compared to version 2.11, this release contains 65575 additional lines of code and it’s the first release to contain more than 320k lines of code in total (as counted by sloccount).

This release is mainly a maintenance and bugfix release, but we have some exciting new features as well:

• Since our group was renamed to “Integrative Transcriptomics“, we have a new splash screen and a new self-signed certificate is used to sign our webstart JAR files.
• Günter Jäger contributed a new version of the Timeseries Alignment Plugin. It allows to compare several timeseries datasets in a three-dimensional visualization embedded into the existing timeseries alignment GUI.
• Matthias Munz wrote a plugin to create an interactive HTML standalone visualization including box plot, profile plot, heatmap, as well as tabular views. The standalone is a single HTML file that can be displayed using any modern web browser. Standalones can also be sent to collaborators by email or placed on a web server to facilitate scientific discussions.
• As a major addition, Stephan Symons introduces MGV, a generic graph viewer for Mayday. It can display graphs enriched with probe and meta information. Graphs can be created from many sources, including Mayday data, biological pathways (BioPax, KEGG), biochemical models and general-purpose graph formats (GML, DOT, GraphML). Both specialized and general graph layout algorithms can be applied. The entire display of graphs, including nodes, edges, node labels, graph title is highly customizable: For the display of data associated with nodes, more than 20 renderers are available, which can be extended by so-called decorators that add meta information. Edges can be drawn in various styles, strokes and colors. Working with graphs is made easy using more than 40 extension plugins for arranging, introducing and evaluating edges, adding nodes, filtering nodes, summarizing and merging nodes. MGV graphs can contain probes from more than one dataset. This allows, among other features, graphical clustering comparisons. For biochemical pathways, all of transcriptomics, proteomics and metabolomics data can be displayed in the same plot. Given mappings of probes and experiments, statistical measures can be applied to perform cross-study comparisons. Included is a Correlation Heatmap that displays the correlation of probes from two datasets.

Here’s a list of other important changes.

Visualization
- New heatmap column type: Expression bars (for black and white printing)
- New heatmap column type: Gene models (if exon data is available)
- Heatmap column headers are now horizontally placed if space permits it
- The "Windows" menu now directly contains the list of open windows.
- Plots can now support selections on experiments (e.g. the tree visualizer)
- Plots can now sort experiments by meta information (e.g. in the heatmap)
- The histogram now is about 100 times faster when binning data to few values
- If colors are assigned by categorical meta-data, an extra window shows the color assignment

User Interface and other improvements
- Drag support: ProbeLists can be dragged out of Mayday (e.g. into excel)
- Drop support: List of probe names can be dragged into Mayday to create ProbeLists
- MIGroup names are now shortened if too long
- Selected elements in tables are remembered after sorting
- The dataset properties dialog is now structured to give a better overview
- New plugin: Summarize experiment on display names
- The Hierarchical Clustering Plugin now supports stored and default settings
- GSEA result probelists now contain a better annotation

Meta Information
- Experiments can be annotated with meta information
- Experiments can be associated with display names for nicer plots
- Drag&Drop support: Meta Information can be dragged into Mayday

SeaSight (Mayday's raw import and normalization tool)
- Memory mapping now uses fewer, larger files
- Memory mapping footprint is reduced by compression if the data permits it
- New locus transformation: Split complex coordinates into their atoms
- Median Polish is much faster now due to synchronized name caches
- CEL import uses more aggressive caching for speed up
- Added "Locus Consistency Check" to prevent disappointment if annotation and mapped
  reads don't agree, e.g. on chromosome names
- Fixed a large bug that potentially concerned all matrix operations from the GUI
- New read summary method: Arcsin
- New read summary method: RPM (the previous version did not work correctly)
- "Naive read count" now offers an option to only use completely overlapping reads
- Great improvements in the SAM/BAM importer in terms of speed and safety
- Experiment properties now also show the initial state next to the current state
- SeaSight can now handly NaN in array expression experiments

New features for developers
- Improved vector subset creation
- DoubleVectors can now be directly created from Double[] as well as double[] arrays
- Drag&Drop support can be extended via plugins
- Memory-mapped structures for booleans
- Memory-mapped structures for genetic strands
- ProbeListManagerView is now a pluggable object, new views can be implemented in the future
- The "Plot" menu of visualizations can now be extended via plugins

Bugfixes and other important changes
- Sequence loading in Genome Browser is much faster now
- GeneMining can now be called on only 2 experiments
- Added detection for first-time use: Mayday now asks for plugin path setting.
  This should eliminate 50% of all support requests
- Removed a dangerous bug in our quantile function
- Time Point scaling and Global Min/Max scaling now work together correctly in Profile Plot
- Time-series alignment now also works with less than 4 experiments.
- FDR correction now also works if some pValues are NaN