CURRENT VERSION OF EASEQ / RELEASE DATE:
V. 1.2 / 2023 03 28
NEW FEATURES (Some are available in the beta test panel)
- Major change in the handling of language and regional settings. EaSeq 1.2 (and future versions) is now ‘culture-neutral’ – it means that some culture dependent features are handled differently for this version. This has several effects, but most likely users will notice that exported values use . (dot) as a decimal separator instead of , (comma). Previous versions allowed users to use local settings e.g. for decimal points, which was more convenient when importing / exporting data. This could however result in unanticipated consequences for the handling of certain double values such as infinity. Peak-calling is now improved compared to previous versions, where certain high-fidelity peaks were lost. See details in the bug fixes below.
- New tools to easily calculate Mean, Standard Deviation (SD), Coefficience of Variation (CV), and log2 fold difference from the mean of parameters, e.g. when signal is quantified from multiple conditions. This is added to the Calculate-tool.
- Export of Datasets now has the option to adapt output to the chromosome sizes of the reference genome (imported from a text file with names and sizes). Reads / coordinates outside of the boundaries will not be exported, this will make output where reads have been extended to strands more compatible with the UCSC genome browser that does not accept coordinates outside of the reference genome.
- Better handling (though not perfect) of screen scaling in Windows.
- Tracks of regions (Regiontrack in the beta-testing panel)
- Calculate tool updated with tabs to encompass more options.
- Export tool updated to allow for filtering of reads outside a user-provided size for each chromosome.
- Major change in the handling of language and regional settings provided by the Windows environment. The previous application of these settings resulted in unforeseens situations where peak-calling depended on the settings. Besides peak-calling, this could affect gating and sorting of regions. Detailed explanation for the effect on peak-calling: For FDR values below the smallest number that the double precision floating point values can handle (app. 10e-308), values are set to 0. When the algorithm calculated –log10FDR values, these values were translated to infinity, which the system can operate as a numerical value. During the processing this was converted to a textstring as it becomes a parameter in the regionset, and then converted back when comparing to the FDR cutoff. On European machines using “,” as a decimal separator infinity is written as “INF” in this textstring, and “INF” was therefore written manually in the strings when the FDR value was 0. However, for unknown reasons, on English (and likely other) machines using “.” as a decimal separator infinity is written as “Infinity”. So when the string was converted back to a numerical value on “.”-based machines, it was not recognized as infinity and is assigned a 0, whereas “,”-based machines recognized it properly. This is regrettable, but not predictable, and no documentation exist (from Microsoft or others as far as I know) on how that infinity is annotated differently in different cultural settings. So this was identified through intense troubleshooting, trial and error, and some users who were kind to provide the feedback that led to the discovery (thank you!).
EARLIER VERSIONS OF EASEQ / RELEASE DATES:
V. 1.12 / 2019 06 26
Contains a minor improvements compared to 1.111.
V. 1.111 / 2019 03 28
Contains a minor bug fix that increases stability compared to 1.11.
V. 1.11 / 2019 03 28
NEW FEATURES (Some are available in the beta test panel)
- Tracks, heatmaps and sorted maps can now visualize data in a strand specific manner (beta, please feel free to help validating this)
- Import of settings optional (can be set in options)
- Checking of consistency optional (can be set in options)
- Reference genome names updated
- Improvements to the tool ExtractSequences (beta tool), including optional minus strand output
- Atypical variable step wig-files are better supported when importing
- When working with multiple figures, then Ctrl+pageup or Ctrl+pagedown changes current figure
- Option to postpose plotrendering in newly imported sessions until figure is displayed. This reduces the workload just after importing a session.
- Stop button added in locusCrawler (beta tool)
- The positions of various panels can be reset from options
- Obs/exp plots not working properly after the introduction of multiple figures. This is fixed now.
- Sort order not handled correctly in Quantify – sorted data sets not selected properly
- Error handling improved for peak-calling and clustering
V. 1.101 / 2018 11 27
- Fixed a new bug introduced in v.1.10 where the regionset panel enters error mode and EaSeq flash repetitive error-messages. This could be triggered e.g. when selecting a subpopulation of regions within a scatter plot.
- The Dataset-, Geneset-, and Regionset-panels are made more error resilient.
V. 1.10 / 2018 11 21
NEW FEATURES (Available in the beta test panel)
- EaSeq now allows the user to make multiple figures with different plots.
- Small vignettes of all figures can be previewed by moving the mouse over the figure icons in the lower part of the interface.
- Navigating the genome or subselecting a set of regions will only affect plots in the active figure.
- All plots in the active figure can be cloned into a new figure.
- Plots can be sent to the next or previous figure.
- The new ‘Multitrack’ plot type allows the signal from multiple datasets to be shown as a density plot. Work on one or many loci. Clicking on the plot will toggle the dataset names from back to front layer.
- Bedgraph files better recognized when importing.
- Dataset import forms offer the user to apply settings to all imported datasets. This requires that datasets do not need to be uncompressed while importing.
- Highlight is disabled for for average plots.
- Import of settings from loaded session files is made optional (change this in Options).
- Checking the consistency of imported data is made optional (change this in Options).
- The sort order of datasets is now handled better in the Quantify tool.
- Dataset and Regionset names are now displayed correctly (taking sort order in to account) in the small text field when putting the mouse over Datasets or Regionsets in the ‘Select sources’ form.
V. 1.05 / 2017 07 14
NEW FEATURES (Available in the beta test panel)
- Option to export underlying pile ups and matrices as text files
- New plot type: Heatmap of regions
- New plot type: Sorted map of regions
- Tool to copy parameters and their values from one regionset to an identical one. Allows time consuming analyses to be carried out in parallel on duplicated Regionsets and then copied back to the original
- Tool to replace text strings in the Regionset parameters values
- Plot settings for XYZ plots: Autoscaling for lower count limit is now optional
- The check of consistency chromosomal naming at data import can now be disabled under options
- No longer fails to load dataset only having reads mapped to one strand
- Minor changes and fixes to user interface
V. 1.041 / 2017 02 21
- Error when importing Bam-files is now fixed.
- Fixed a potential error when exporting pdfs.
V. 1.04 / 2017 01 27
- Figures can now be exported in Pdf format, which makes it simpler to later change e.g. axis names in e.g. illustrator.
- App. 40 new Color schemes has been added.
- Split reads in Bam files are now randomly assigned to one of the areas that it covers. Probabilities depend on the degree of overlap between the read and each area. This is an improvement on previous behavior where. E.g. split RNA-seq reads at intron/exon junctions were mapped to the exon with the lowest coordinate.
- The Get Sequences tool (in the Beta-test panel) now optionally provides Ns at repeats.
- EaSeq will check for write access to the home-directory at startup.
- Intrapolation mode in the sortMap and NSMap plot types is now adjustable.
- The Colocalization tool now offers the option to measure distances by borders and find regions with least distance from border to border (and thereby largest overlap) instead of the regions with the lowest distance from center to center.
- Improved handling of non-numerical and extreme values in 2d histograms. Non-numerical values and the logarithm of negative and zero values are now optionally displayed in lowest bin, when the ‘include all’ (formerly known as ‘include outliers’) checkbox is checked in plot settings.
- The Calculate tool now also offers calculation of the size of each region.
- The Quantify values tool now can use the normalization type entered in the describe dataset (e.g. custom factor or reads pr. million).
- The Skip button has been removed fromt eh waiting form, when it was not functional.
- Users are notified of the space and time requirements defore download of the demo session.
- The algorithm checking for consistent and conventional chromosome naming can now be instructed not to mention inconsistent chromosome names multiple times when non-standard names are used.
- Minor updates to the user interface for the Squeaks.
- The interface in Cluster regions by dataset is slightly modified.
- Hints are now available when clicking the button, also after they have been deactivated.
- Plots: values for axises are now loaded correctly from a session on a machine that use a different decimal sign than the one saving the session.
- Dataset normalization factors are now loaded correctly from a session on a machine that use a different decimal sign than the one saving the session.
- Improved handling of headers and comments in imported files.
- Minor change in autogenerated description from the colocalization tool and for plots.
- Annotation tracks are now updating when a new locus on clicked in a heatmap. That was not always the case before.
- The order of samples in the selection overlay has been corrected from erroneously being reversed in box, line, bar and density plots.
- Some inaccuracies when gating from plots has been corrected
- The Extract tool has been corrected, so it now actually takes the values entered for the start offset in to account. Previous version ignored this.
- When a region parameter is deleted, it is now only the plots using that particular regionset that have their settings changed.
V. 1.03 / 2016 07 07
- New Beta-testing panel. Found in the right side of the labdesk – close to e.g. the navigation panel.
- The new ‘Locus Crawler’-tool allows users to automatically export genome browser tracks for a large number of loci or genes (found in the beta-testing panel).
- The new ‘Export sequences’-tool allows users to download and export the sequence corresponding to a set of regions. Useful for motif analyses e.g. using the MEME-suite.
- More flexibility when handling decimal signs of imported coverage values from bed-graph files.
- Error at startup during first run fixed.
- Strand orientation in Quantify-tool is now visualized correctly. For regions on the negative strand a skewed up the visualization in the window – The actual quantified values are were not affected and correct all along.
- Errors in the autogenerated descriptions from clustering are now corrected: 1) position was wrongly reported as ‘Start’ when selecting ‘Fixed size’ and ‘Center’ when selecting relative size instead of getting its values from the selected positions, 2) Log-transformation was not listed, 3) number of bins was always reported as 0, and 4) data were always reported as not normalized. The actual clustering was unaffected by this.
- Newly gated regionsets are now displayed in the list. An error kept them from being displayed.
- Improved errorhanding when Regionset processing fails.
- Minor changes to the behaviour of hints.
V. 1.02 / 2016 06 01
- Headers in bed-files and bedgraph files are now supported better.
- EaSeq gives a warning if e.g. datasets are loaded as regionsets or vice versa.
- Chromosome names and cases are not changed unless the user instructs it.
- The nearest region is now selected when no regions are selected in Histograms, scatter plots etc.
- The number of regions is now included on histogram title.
- Slightly skewed histograms (1 bin) on the X-axis are now corrected.
- Icons for panels updated, now the number of new chat-messages is shown (squeaks).
- Error occuring at midnight when sending usage statistics is corrected.
- Panel repositioning error fixed.
- Panel positions are saved when exiting EaSeq and in sessions.
- Minor changes in interface and tutorial.
V. 1.01 / 2016 04 15
- Wig files offers an option to expand span when loading.
- Spans in wig and bedgraph can now be of any length.
- Values in exported wig and bedgraph files are no longer rounded to integers.
- Fixed bug when importing wig files with negative coordinates.
- Fixed bug wrongly stating that the number of chromosomes had exceeded 100 upon import.
- Various panels can now be dragged around – quick fix for a bug where the navigation panel is below the border of the labdesk. Interface will get a larger overhaul later.
- Add new reference genomes to ‘download geneset’ dropdown box.
- Histogram default size is changed.
- Show information about the content the mouse hovers over when selecting datasets / regionsets / parameters for new plots.
- More environment variables are cleared when a new session is loaded.
- Change the autogenerated text for the Phantom and Peak tools to add a paper to cite for the original development of the Phantom Peak coefficient.
- Text updated in Gate, Sort and Quantitation tools and when an unsorted bed or a bam file is imported.
- Links to online content in about.
- Chromosome renaming function is optimized and its usability improved.
V. 1.0 / 2016 02 29
- Reference and links to our paper are added.
- Chromosome renaming function is optimized and its usability improved.
V. 0.992 / 2016 02 16
- Varying or unconventional naming of chromosomes is detected when new data are imported, and users can rename them after import.
- Error messages are more informative when columns are undefined during import.
- Several minor changes to data import to make it more robust.
V. 0.991 / 2016 02 09
- Regionsets without coordinates can now be imported by finding them in a geneset based on gene names or accession numbers
- Loading genesets without coordinates has been corrected and made case insensitive
- Changed cases for chromosome names and case sensitivity lead to e.g. chrX not being recognized when annotating. Gcoloc, Annotate, Coloc, Controls, and Modify/Merge: Chromosome names are now made case-insensitive. Case has been updated for several tools.
- Error messages when importing has been updated
- Formula example in the calculate tool has been updated to show the real log-base when log-transforming numbers
- The label showing that a newer version of EaSeq is available is updated to link to the current download page
- Closing EaSeq can be cancelled
V. 0.99 / 2016 01 22
- EaSeq will offer to add ‘chr’ to chromosome names without when loading Datasets, Regionsets, or Genesets
- Log-mode button was enabled without having any effect in plot settings in Heatmaps and Ratiometric heatmaps. Now, it is properly disabled.
- Regionsets no longer hang after annotating them with gene names
- The pattern search tool is fixed so it no longer crashes
V. 0.984 / 2016 01 19
- The performance is optimized when moving the mouse over a plot
- The Inspector now shows descriptions 1000 fold faster when clicking on a plot or dataset
- Position viewer does not go in error mode when showing regionset with 0 parameters
- Descriptions are now associated to each new parameter when colocalizing, not only the first one
- Three new parameters are added when colocalizing. They contain coordinates for the closest regions in a colocalization analysis
V. 0.983 / 2015 11 13
- An error in the size calculation of Regionset overlap analysis has been corrected
- Venn diagrams from Regionset overlap analysis includes numbers for bp of each set of regions and an updated legend
- Wig-files without header can be loaded and the header composition can be more flexible
- Now only unfrozen, sorted plots showing the sorted Regionset are updated
- Sort order is now conveyed correctly to plots when a new Regionset is selected in plot settings
- The region that is shown in the position viewer now takes changes in the sort order into account
- Z-scatters now have the checkbox for log-count enabled
- Annotation is no longer lost when loading from a session file
- Annotation tracks no longer shows error when loaded from a session file
- The autoscale count setting in plots is now loaded correctly from a session file
- Elements generated from overlap analysis now no longer show errors when loaded from a session file
- Obs/Exp plots no longer reports errors when redrawn
- Deleting a plot no longer accidentally leads to deletion of other plots
- Plots no longer disappears when moving multiple plots
- Plots from Regionset overlap analyses are now positioned better on the labdesk
- Duplicated plots are now positioned better on the labdesk
V. 0.982 / 2015 10 07
- EaSeq no longer crashes when regions are sorted while heatmaps display them
- The ‘Autoscale max count’ setting for plots is now saved correcly
- The ‘Max count’ setting for plots is now loaded correcly
- EaSeq checks for newer versions both at midnight and startup to ensure that new versions are announced
V. 0.981 / 2015 10 05
- Write access to the folder is no longer required when importing bam-files
V. 0.98 / 2015 10 02
- A bug that affected Bam-file import was fixed
- Tutorials were tested and had multiple fixes
- A bug in parameter-based clustering was fixed
- A bug making it impossible to exit from a note when changing its text was fixed
- Keyboard shortcuts are now deactivated when settings are shown for plots
V. 0.97 / 2015 09 30
- Hints were added to explain concepts and guide the user when using tools
- Relevant online information is linked from hints
- Tools that produce output similar to that of popular Linux tools (e.g. bedtools) now have hints explaining the similarity
- Tutorials were updated
- Relevant tutorials now have hyperlinks in the Demo
- Various parts of the tutorials were fixed – others remain to be crash-tested and fixed
- Positionview now updates when regionset is clicked
- K-means clustering based on parameters adds a bar showing the clusters automatically
- Overlays can now optionally be included in snapshots
- Heatmaps now also have overlays
- All plots update when settings or order are changed
- Heatmaps etc. can show the sort order as an optional overlay
- An annotation map has been added to illustrates sort order next to heat maps
- Annotation tracks and maps now docks at a ‘master-plot’ to ensure consistency in axes and values
- The plot user interface had a complete and thorough makeover
- Better positioning of plot buttons, automatic overlays with relevant information, and more smooth interaction
- More detailed error reporting in plots that fails
- Descriptions of plots are updated to include how other regions within the window are depicted
- The colocalization tool now supports calculating distances to adjacent regions within a regionset
- Plot producing tools e.g. co-occurrence and overlap are updated to make plots early on
- Normalization tool interface redesigned
- Users are notified when parameters or new data are added
- Progress bar colors changed to blue during progressing work
V. 0.93 / 2015 07 08
- Datasets now support custom normalization values.
- The Describe Dataset form has been updated to support custom normalization.
- Plots are updated to support custom normalization as specified in the Describe Dataset form.
- The calculator can now also handle calculations that only involve a single parameter.
- The Export plot data button was removed until data export is supported by all plot types.
- An Add notes button was included.
- The Region-viewer was rebuilt from scratch, simplifying parameter renaming and deletion.
- Most other plot types now highlight the data point that correspond to a selected rows in the region viewer.
- The Dataset-, Regionset- and Region-viewers now autoupdate to show the corresponding region when a selection is made in a plot.
- The orientation of a single selected region in the Region-viewer is now correctly identified for coordinates on the negative strand.
- Increased consistency in the start and end coordinates used for plots, navigation, and the region viewer.
- Data in session files are now recognized correctly when sessions are shared between users from countries with different regional and language settings in Windows.
v. 0.92 / 2015 07 01
- EaSeq is no longer time-limited and does not expire
- Minor changes to the main interface
- The Bam-file import code is updated to support files exported from bwa
- The Bam-file import form is updated with standard settings for ChIP-seq and RNA-seq
- The Inspector no longer overwrites the name of a renamed dataset / regionset with that listed in the inspector
- The Inspector no longer overwrites parameter names with the previous name
- Exported Regionsets are now ordered according to the latest sorting / clustering
v. 0.91 / 2015 06 17
- First public test version