Date | Updates |
2014-02-06
version 1.08 |
This update fixes a serious bug discovered in earlier versions related to
local caching of downloaded Region data tracks. (Read more)
|
2013-04-24
version 1.07 |
Support for some of the tools
from MotifSuite
led to the discovery of a bug in v1.0 of MotifLab which was subsequently
fixed in v1.05. However, this bugfix actually resulted in problems with
the ChIPMunk
program. With this new version 1.07, all programs should hopefully work once more.
|
2013-04-16
version 1.06 |
- Added new statistic to "Compare motif track to numeric track" analysis
which will count the number of motif instances (TFBS) where the average
value of the numeric track within the TFBS is greater than or equal to a
specified threshold value.
- Fixed a bug in the "Compare motif track to numeric track" analysis
which caused the "sum" column to be output instead of the "min" column
when outputting the analysis to HTML or RawData formats (the "sum" column
was hence output twice).
- Added an extra "Help" button to the dialogs for the
"analysis" and "output" operations which will display the HELP page for
the selected analysis or data format respectively (note that the
documentation itself might not be completed yet for every analysis/data format).
- Added functionality to create Sequence Numeric Maps based on statistics
- Added functionality to extract either "assigned entries" or "unassigned"
entries from a Map object. The "assigned entries" are those entries that
have explicitly assigned values in the map, whereas unassigned entries do
not have explicitly assigned values and will thus return the default value.
- Changed the MapFormat and MapExpression data formats so that the
"Include entries" parameter now refers to a Collection containing the
entries to include in the output.
|
2013-04-09
version 1.05 |
- Added "Center" sequence alignment anchor to "Motif position
distribution" analysis.
- Added "Center" sequence alignment anchor to Positional Distribution
Viewer tool and added possibility to only base the histgram on a
(sub)collection of sequences. (Also fixed some additional minor
problems with this tool).
- Fixed some layout issues in "Configure Data Tracks" dialog
and also included functionality to add new tracks and data sources from
an XML-based configuration file (or by directly copy/pasting or editing
XML-formatted configuration entries in a simple text editor)
- Made several additions to the EvidenceGFF data format:
- Added a "skip standard fields" option
which will omit the standard GFF fields and only output the
evidence column(s)
- Added new "sequence" evidence codes to output different
sequence properties (e.g. sequence name, gene name, sequence
location, organism and genome build, etc.)
- Added a "text" evidence code to output literal text.
- Added a "percentage overlapping" code which
will output how large fraction of a TFBS is overlapped by a
region from another track.
- Added MapExpression data format for (numeric) Maps which can
output maps in arbitrary user defined formats and input maps
based on user defined regular expressions.
- Added "Include/Exclude Marked Sequences" options to the
context menu in the Sequence Collection dialog
- Fixed a serious bug which caused MotifLab to crash when
"Execute Current Line" was selected in the Protocol menu.
- Added option to use only a (sub)collection of the sequences
when creating a new Background Model based on a DNA track
- Fixed problems when importing motif collections in TRANSFAC format
- Fixed rendering problems in Motifs Panel when motifs are updated
- Fixed a minor bug for the draw tool so that regions can now be deleted by pointing
at them with the mouse and pressing the DELETE key (previously
this worked for all tools except the draw tool).
- Fixed a problem with parsing the condition for the "interpolate"
operation and also changed the command for this operation somewhat
- Added option to draw borders around the regions in region
tracks. These borders can either be in black or in a darker shade of the color
of the region. To enable borders for a specific track, select the chosen
border style from the context menu in the Features Panel or select
the track in the panel and press the "B" key to cycle between the
three different styles (no border, darker color or black).
The current implementation of the border is not very nice in
close-up (since the width of the border increases as you zoom in),
but I will try to improve this in the future.
- Added "Info" button to External Programs Repository dialog
which makes it possible to display information about a selected
program in the repository in a popup dialog without actually
installing the program first.
- Added support for several of the programs
from MotifSuite,
including MotifLocator, MotifSampler, MotifComparison and CreateBackgroundModel.
The configuration files needed are available in the repository. Most
of these programs will also work with earlier versions of
MotifLab. However, use of positional priors in MotifSampler
requires this new version to function properly.
|
2012-12-20
version 1.0 |
Official release of version 1.0
|
2012-12-17
version 0.9999 |
- Fixed bug in "rank" functionality for Modules and Sequences (the
rank operation would previously always return Motif Numeric Maps rather than maps for
the correct type)
- Fixed bug regarding "span" statements in WIG data format.
- Changed command syntax for "count" operation
- Added possibility to use literal constant as 'anchor' position for
"distance" operation
- For sequences specified in BED format, we have added the
possibility to base the sequence names on the coordinates
(chr_start_end) rather than assigning them default incremental names (SequenceX)
- Added even more "extract" options, such as extracting "top/bottom X"
entries from maps as collections, extracting "size" from collections
and to extract random subcollections.
- Plus additional minor improvements and bugfixes.
|
2012-11-23
version 0.9998 |
Just some minor additions.
Added more "extract" options for Collections and Numeric Maps.
Added optional "flanks" for base cursor (set in sequence window context menu).
|
2012-11-15
version 0.9997 |
- Added ScerTF as
a predefined Motif Collection
- Added possibility to specify "preferred mirror" servers that
will be used instead of the original server specified in the data
sources. This is very useful if you have created your own mirror
of the UCSC Genome Browser (or just want to use a mirror with
less traffic). See the "tips & tricks" section for how to do this.
- Added "-config" (or "-c" for short) option to the CLI-client that can be used
to set certain MotifLab configuration settings without going
through the Options dialog in the GUI.
- Minor bugfixes and GUI improvements
|
2012-10-26
version 0.9996 |
- Fixed bug in SimpleModuleScanner
- Added tool to extend sequences
- Added new keyboard short-cut "U" to Features Panel which will sort
the sequences in a region track according to the number of visible
regions in each sequence (descending)
- Fixed a very embarrassing bug which led to bundled motif
discovery/scanning programs being unusable the first time MotifLab was
started. (After that they would work normal.)
- Improved the performance of the bundled motif scanning program
'SimpleScanner' by using multi-threading to take advantage of multiple CPU cores.
- Added new bundled motif scanning program 'ConsensusScanner' which
searches for matches to IUPAC representations of motifs (rather than
using the PWM representation).
- Changed the 'Include' output-parameter used by many analyses to
be based on user-defined collections rather than predefined
criteria.
- A lot of other smaller bugfixes and GUI improvements
|
2012-10-01
version 0.9995 |
- Finally added support for compound conditions (multiple conditions
connected by the boolean operators AND/OR). The conditions can be
grouped and nested to arbitrary levels.
- Added "motif similarity" analysis which compares a single motif to a
collection of motifs using all available comparison metrics. Also made
a distinction beetween "similarity metrics" (which give higher scores to
more similar motifs) and "difference/distance metrics" (which give lower
scores to more similar motifs).
- Improved the GUI appearance of Benchmark analyses with sortable
tables. Added a few more performance metrics and the possibility to
specify the minimum required overlap in order to call a prediction "true
positive" on the site-level. Also improved calculation of Matthews
Correlation Coefficient (nCC) which could previously sometimes fail due to
precision-problems in some extreme cases.
- Results for single operations that are executed interactively in
the GUI are now shown immediatly in popup dialogs when appropriate. (Previously, only the
results for the "analyze" operation appeared in this way.) This
behavior can be suppressed by holding down the SHIFT-key when clicking
the OK-button in the operation-dialog.
- Added "max distance" parameter to "interpolate" operation
- Fixed bugs affecting training/validation dataset sampling for
PriorsGenerators
- Fixed bug affecting external programs that output results to STDOUT
rather than writing the results to files
- Updated the protocol language for a few operations in order to make
the commands seem closer to natural english language
- Additional minor bugfixes and GUI improvements
|
2012-08-29
version 0.9994 |
Fixed a few bugs (including some serious ones affecting execution of
external programs that were introduced in the previous release).
|
2012-08-13
version 0.9993 |
- Added support for ChIPMunk motif
discovery tool
- Improved support for user-defined properties for regions in conditions
and arithmetic operations.
- Added "BED" tab in Sequence Dialog which allows sequences to be specified
as genomic regions in BED-format.
- Updated the "score" operation so that it can now score
direct/reverse/relative/opposite strand in addition to the best score from
both strands.
- Added "color boxes" as optional first column in HTML-output for all motif-
and module-analyses
- Added new type of evidence to EvidenceGFF-format which outputs the
distance from the target region to some other region in a track.
The syntax "<regiontrack>=distance to closest" (or
distance to 'any')
will output the distance to the nearest region in the specified region
track. "<regiontrack>=distance to <Selected>"
will output the distance to the closest region whose type is listed in the
'Selected' data object (which can be a Collection or Text Variable).
"<regiontrack>=distance to interaction partner" will
output the distance to the nearest interaction partner in the given
track. It is possible to use the 'non-overlapping' qualifier with either of
these.
- Fixed a +/-1 mistake in the "merge" operation
|
2012-07-02
version 0.9992 |
- Added "Favorites" dialog to the Data menu where users can register often
used data items for easy access.
- Added new 4-field input/output format to "Location" data format for
sequences. Also added possibily to specify custom output format.
- Added new output data format for modules: HTML_ModuleTable.
- Added optional drop-shadow effect for expanded region tracks in close-up
view. This is blatant eye candy and is disabled by default. To toggle this
setting, select the track in the Features Panel and press the W-key.
- Added functionality to select the whole sequence or invert the selection
windows in a sequence. This functionality is available from the sequence
context menu or by pressing keys A or I.
- Minor bugfixes and improvements to the GUI.
|
2012-05-18
version 0.999 |
- Added a new configuration dialog which allows users to configure which
organisms and gene identifiers MotifLab will recognize.
- Added possibility to edit the "startup script" which is executed when
MotifLab's GUI is started (from the Configuration menu)
- Added "sort/group" buttons to the Features Panel and Data Objects Panel.
These buttons link to functionality to sort data objects alphabetically and/or
group them by data type.
- Improved the "Benchmark" analysis by adding site-level statistics for
sensitivity, PPV and ASP.
Also, switched to using a "format" string to specify which statistics to
include in the output
(and their order) rather than using a bunch of checkboxes.
- Improved the wildcard functionality available for some display settings by
allowing wildcards to be used in
combination with a prefix or suffix rather than just a single wildcard.
- Added a tooltip to the logo column of the "compare motifs panel" which
shows the pairwise alignment of the
selected motif and the target motif.
- Added a button to the Motif Score Filter which can be used to specify
whether the filter should hide regions with
scores ABOVE (or equal to) the selected cutoff value or BELOW the
value.
- Improved the Crop Tool by allowing Numeric Variables or Maps to be used to
specify the number of bases to crop off the start/end of sequences.
- Fixed bug in GTF data format for attributes containing spaces and added
support for referencing
user-defined properties in the alternative "format" string for GFF data
format.
- Additional bugfixes and GUI improvements.
|
2012-04-25
version 0.998 |
- Drastically improved performance when saving sessions (especially when
saving to 'slow' devices such as network discs).
- Fixed a somewhat serious problem in regard to using external
programs (affecting users of "Weeder")
- Minor bugfixes and improvements to the GUI
|
2012-04-16
version 0.997 |
- New operation: "rank". Can be used to extract the rank order of
motifs,
modules or sequences based on a Numeric Map, a numeric column in an
Analysis or a numeric property of
the data objects themselves. It is also possible derive a combined rank
order based on the (weighted) rank sum of several properties.
- Added two new motif comparison metrics based on symmetric
Kullback-Leibler divergence and Chi-squared test.
- Minor bugfixes
|
2012-03-23
version 0.996 |
This version contains quite a number of major updates and corrections and
some of the new functionality is no longer compatible with earlier version
(this can also affect saved sessions). If you need to use the previous
version (v0.995) you can find it here.
- New operation: "copy". Can be used to create identical copies of any data
object
- New operation: "difference". Can be used to compare two data objects and
highlight the differences between them.
- New operation: "collate". Can be used to combine data-columns from several
different analyses into a single "meta" analysis.
- New operation: "discriminate". Can be used to convert a "regular" priors
track into a "discriminative prior".
- New operation: "prune". Can be used to remove duplicate regions or similar
overlapping regions from Region Datasets.
- New analysis: "Region Dataset Coverage". Can be used to assess how large
portion of each sequence is covered by the regions of a Region Dataset.
- New analyses: "Count Module Occurrences","Count Region Occurrences" and
"Compare Region Occurrences" which works the same way as the old "Count
Motif Occurrences" and "Compare Motif Occurrences" except that they work for
modules and general regions instead of motifs.
- Replaced StringVariable and StringCollection with a single new
TextVariable data type
- Added possibility to reuse images created for motif and module logos
across multiple analysis when output to HTML.
- Added support for Cascading Style Sheets for HTML-output and JavaScript to
sort HTML-tables. (this only works when viewed in external browsers).
- Added "type-replace" transform which can be used to replace the region
type of region based on regular expressions.
- Added possibility to create random collections with size specified in
either absolute or relative terms (undocumented feature only available in
protocol).
- Added several new "display setting" statements including "restoreSession",
"saveOutput" and the general "setting".
- Added optional "-saveSession" and "-inputSession" arguments to
CLI-client
- Added possibility to display several independent histograms on top of each
other in the Positional Distribution Viewer tool, and also added auto-update
functionality so that histograms are updated if the visibility of regions changes.
- Added "within region" parameter to "Count Motif Occurrences"
- Added pie-chart to "Compare Region Datasets" analysis
- Added possibility to specify default programs to use for motif/module
scanning/discovery.
- Made context menus and keyboard shortcuts more consistent throughout.
- Grouped together operations in context menus in the data panels and made it possible to
select applicable analyses through context menus.
- Added option to mark parameters as being "advanced". Advanced parameters
are hidden by default in parameter panels.
- Added possibility to assign values to, and change values of, custom region
properties throught the use of arithmetic operations.
- Added possibility to control the appearance of bar-charts (colors,
borders, shadows, gradient fill). This is so far an undocumented
feature.
- Added possibility to change the colors used by the protocol editor.
- Added option to sort sequences based on selected criteria to the context
menu for feature datasets. (This sort of works as a short-cut to the
functionality in the Sort Tool).
- Added functionality to create a Numeric Variable based on the size of a
collection: X = new Numeric Variable(somecollection)
- Improved the SimpleModuleSearcher algorithm with several new options
- Added functionality to create Motif Partitions based on known alternatives
for each motif (creating "equivalence clusters" of similar motifs)
- Added functionality to automatically create Modules based on known motif
interactions.
- Added functionality to create Sequence Collections based on a chosen
statistic, e.g. which sequences contains a given motif(s), or which
sequences have GC-content>50%.
- Added proper support for the FIMO motif-scanning algorithm.
- Fixed problems with UNC-paths when executing external programs.
- Improved SimpleModuleScanner so that it now can use Module Collections as
input rather than a single Module. (It is still rather slow for many
modules, though).
- Added functionality to create and display histograms in Motif Position
Distribution Analysis.
- Added column header tooltips for most tables with numeric data which shows
information about the values in the columns, such as min, max and average
values and also a histogram.
- Added possibility to have "horizontal gradient fill" for regions
- Added functionality to extract the rank-order of values in a Numeric Map
as a new Numeric Map.
- Added new output-format for motif tracks: MATCH
- Added a few more "clear data" options to the top-level menu in the
GUI.
|
2011-12-21
version 0.995 |
- Implemented support for making concurrent download request
to data servers so that data for several sequences can be downloaded in
parallel rather than just one at a time. This can significantly speed up
the download time for larger datasets. However, since it also places
additional strain on the data servers, this option is disabled by
default.
- Implemented functionality to allow new Priors Generator classifiers
to be trained 'on-the-fly' in a protocol script by specifying the setup
of the classifier and training process in a configuration file.
- Fixed an embarrasing bug in many Motif-dataformats which lead
files containing only a single motif to be unparsable
- Made it possible to specify the preferred order of 'mirror
datasources' (different sources for the same feature+genomebuild) by
reordering sources in the 'DataTrack Configuration'-dialog through
drag'n'drop.
- Added a new type of transform (to the transform operation) called
'odds' which is the same as 'logit' but without the log: F(p)=p/(1-p)
- Added possibility to 'extract' a list of the types of regions that
are present in a Region Dataset as a String Collection
- Added support for TSS-relative coordinates in GFF and
EvidenceGFF data formats
- Added option to place TSS at either position '0' or '+1'
(default). If the
latter option is selected, a TSS-relative coordinate system will skip
the 0-position and proceed directly from position -1 to +1.
- Added option to show cumulative histogram instead of normal
histogram for the 'Numeric Dataset Distribution' analysis
- Added possibility to "skip unrecognized gene IDs" in the Sequence
Dialog
- Added '-retry off|prompt|<seconds>' option to the CLI-client which
specifies how the client should react when MotifLab encounters
errors.
- Plus several minor improvements and bugfixes
|
2011-11-08
version 0.994 |
Added some minor features and fixed a
few embarrassing bugs.
- Added support for embedded bedGraph lines
within WIG
files.
- Added "Import" button to the "Manual Entry"-tab in Sequences Dialog
so that locations of sequences to analyze can be imported from a
file. (Either 8-field or 10-field formats)
-
- Added support for parsing Motifs in TAMO format (but
not for writing output)
- Improved speed-performance of "show reverse complement" option in
Motif Dialog.
- The range of values for Motif Score Filter is henceforth only based on
visible sequences in visible tracks, and the range will be updated
whenever the visibility of any tracks or sequences are changed.
- Fixed problems with initialization of consensus motif, IC-content
and GC-content from log-odds matrices read from file.
- Fixed Compare
Region Datasets analysis (which did not work at all because of a
newly introduced bug).
- Fixed protocol parse error complaint caused by <= operator in
conditions
- Added support for extracting "top:X" and "bottom:X" entries in a Map
as a Collection, and to extract "_DEFAULT_" value as Numeric
Variable.
(This was actually already possible, but only as a "hidden feature")
- Fixed bug in argument for Transform operation (affecting GUI)
- Fixed bug
in Priority
format so that it is now possible to use SPACE as field separator (this
did not work properly before).
- Allowing double-stranded priors in
Priority
formatted files. However, in these cases only the first half of the values will
be used.
|
2011-10-28
version 0.993 |
- Updated AffinityScanner motif scanning algorithm, so that it can
use either raw or log-likelihood scores (both absolute and relative).
- Made
the Motif_Properties
data format more general and expressive (but perhaps slightly less user-friendly)
|
2011-10-18
version 0.992 |
- Minor bugfixes and improvements to the GUI
- Added "welcome screen" which is shown the first time MotifLab is
started
(Also available under the Help menu)
|
2011-10-06
version 0.991 |
A few more minor updates and improvements
- Improved performance when adding duplicate sequences (which is not
really allowed)
- Added option to skip unrecognized identifiers when entering a list of
gene IDs in Sequence Dialog
- Fixed bug when creating Motif Numeric Map from a track
- Added
new display
settings: regionColor, regionvisible, showRegion, hideRegion, updates and refresh
- Added "verbose mode" to CLI-client
- Added possibility to sort output from "count motif occurrences"
analysis by p-value
|
2011-09-30
version 0.99 |
This release comprises several updates and improvements. Although I plan
to add some additional support for a few external programs in the immediate
future (ModuleSearcher, ClusterBuster and ChIPMunk), with this version
MotifLab is now considered to finally be stable, feature-complete and free of known bugs.
The version number will therefore be upgraded to version 1.0 after a short
period of testing.
- Added "DAS wizard" to Datatracks configuration dialog. The wizard allows
users to easily search and add new Region Datasets found in the DAS-registry (www.dasregistry.org)
- Added dataformat CisML for output-formatting and parsing region datasets (especially motif tracks)
- Added possibility to create Motif Collections based on motif properties
- Added support for motif scanning
tools MATCH
and CLOVER
- Made all arithmetic operations, including transform
and threshold applicable to Expression Profiles and
Numeric Maps
- Added several new display settings statements
- Added option to show/edit motif representation (matrix, consensus and
logo) in reverse complement
- Fixed persistant GUI state for the data side panel (containing
features, motifs and data panels)
- Made context menus for tables showing sequences, motifs and modules
more consistent
- Added support for headers in Expression Profiles
- Added possibility to create Motif Numeric Maps based on motif
comparison scores (via context menu in motif comparison panel)
- Changed the orientation setting called 'From Sequence' to
'Relative'
- Added option to include constituent single motif sites when outputting and
parsing module tracks. This will store all relevant data in the GFF-file so
that module tracks can be completely restored from file.
- Added possibility to return either raw or log-likelihood scores when using
the score operation, and to return either absolute or relative scores.
- Added new motif-comparison function: IC-weighted Pearson's
correlation
- Finished implementing prompts for all data types in CLI-client
- Added support for mirror servers for data sources
- Added possibility to comment/uncomment selected lines in protocols (in
Protocol menu)
- Added option of converting regular region tracks to Module tracks (in
Features panel context menu)
- Attempting automatic conversion of all loaded Region datasets to Motif or
Module tracks if possible
- Added "anchor" setting to Positional Distribution Viewer
- Added possibility to specify anchor and relative start/end coordinates for
each sequence in Sequence Input Dialog
- Added possibility to view and edit additional region properties (by
double-clicking on a region site while holding down CTRL)
- Added possibility to define which colors to use to represent
up/down-expressed sequences in Expression Profiles
- Added possibility to delay resolving references when creating collections
or partitions "From List" (this will keep the original references in the
protocol)
- Finished tooltips for non-expanded module tracks
- Fixed faulty implementations of ROC and Precision/Recall curves
- Hopefully, fixed problems with motif sequence logos across platforms
- Renamed some of the analyses
- Updated descriptions of operations (used for tooltips etc.)
- Fixed several minor bugs and made improvements to the GUI
- Added new Welcome screen which will be shown on first startup (also
available from the Help-menu)
|
2011-08-22
version 0.983 |
- Added support for user-defined properties for motifs
- Updated context menus for tables and panels containing sequences,
motifs and modules to make them more consistent
- Fixed a few syntactic errors in the README.txt
and runMotifLab.bat files that are included in the downloadable
ZIP-file
|
2011-08-15
version 0.982 |
- Added two new condition operators applicable to regions:
weighted
average and weighted sum.
The new operators are alternatives to the earlier 'average' and 'sum'
operators that considered the average value (or sum value) of a numeric
datatrack inside the sequence segment spanned by the region.
The new operators are meant to be used with motif tracks and will
default to the standard non-weighted operators for other region tracks.
In the weighted instances, the numeric track value in each position is
weighted by the (normalized) IC-content of the aligned motif in the
corresponding position. E.g.: Consider a binding site for a motif with
consensus "nnnCACGTGnnn" (size=12bp) which is being compared to a numeric datatrack
with phylogenetic conservation values. In the conservation track the six
middle bases are perfectly conserved while the other bases are not
conserved at all. According to the standard 'average' operator the
average conservation score of this TFBS will be 0.5 since only six of the
twelve bases are conserved. However, if we use the 'weighted average'
the six middle bases will be given more weight than the flanks since the
middle part has higher IC-content. In this example the weighted average
will be closer to 1.0, which signifies that the 'important part' of the
motif is conserved (while the non-important parts are not considered).
- Renamed the old condition operators upstreamValue
and downstreamValue which considered the most upstream (or
downstream) value within a region relative to the orientation of the
sequence to sequenceUpstreamValue
and sequenceDownstreamValue. At the same time, two new operators
were introduced which considers the most upstream (or downstream)
values of the region relative to the orientation of the region
itself. These were then called regionUpstreamValue
and regionDownstreamValue.
- Changed the orientation setting called "From Gene" which is used by
many data formats to "From Sequence" (but the old setting can still be
used to retain backwards compatibility).
|
2011-08-12
version 0.981 |
- Introduced new data types for Module Partitions and Module
Numeric Maps
- Updated the dialog interface for Sequence Collections to be more like
other collections, partitions and numeric maps
- Updated the dialog interface for Modules by introducing more
controls for functionality that was previously only available
through context menus. This will hopefully make the options for
configuring Modules more visible to the user.
- Added a new output format for Region tracks called EvidenceGFF
- Minor bugfixes and improvements to the GUI
|
2011-07-13 | Made first BETA-version of MotifLab
publicly available |