Version-1.5.3 (aka 2004.1) Release Notes, November 2004

The is a general new-features and bug-fix release of the Staden Package. There are no massive changes, but I hope you'll find the updates useful still. Windows users may find a larger difference as the previous release was not build on that platform; please read the 1.4 release notes too.

This release also includes much needed updates to the documentation. It is provided in PDF format; if you wish to obtain the old PostScript format we previously provided please use the Acrobat (or similar) tools.

For the full list of bug fixes and gory details of all files changed, please see the ChangeLog. Here is a summary of new features or improvements.

James

PS. Incase you were wondering, version 1.5.0 to 1.5.2 existed mainly as tagged releases in the CVS tree, but correspond to internal test versions used within the Sanger Institute.

Gap4

• New plot: 2nd-highest confidence. This plot shows next highest confident base per consensus location. (This plots the same as the contig editor "consensus discrepancy" search.)

• New plot: discrepancies. Much like the 2nd-highest confidence, but with some added statistics to compute the likelihood of this position containing two different sequences.

• New command List Base Confidence on View menu. This tabulates base call confidence values along with their agreement or disagreement with the consensus. This is a useful measure for the amount of variations within an assembly (whether they are real SNPs or assembly errors).

• Added a "Join To..." menu to the contig editor names panel. This brings up the join editor based on read-pair linkage between contigs.

• New editor search mode - by consensus discrepancy. Unlike the normal discrepancy search, this finds locations where the consensus scores computed from each base-types in turn indicate a problem.

• Mousewheel support has been added to the Y-scrolling of the contig editor and trace displays.

• New function: "quality clip contig ends" to trim back low quality data off the end of contigs. The purpose is to correct for phrap assemblies where clipping is typically done based on disagreements, and disagreements at the very end when depth=1 do not happen. This leads to 'rubbish' being at the contig end, devaluing the FIJ with "hidden data" turned off.

• The editor reference sequence will now always be shown at the top of the editor.

• When auto-diff traces is enabled, the editor will automatically pick wild-type traces to compare against if none have been set.

• Improved the Y-scaling for auto-diff. It is now beneficial to enable!

• Added a setting in the editor to automatically bring up read-pairs of traces. That is, if a forward strand sequence is clicked on and it overlaps with the reverse strand, then both traces are shown.

• Tags may now contain simple GUIs in place of the text panel. The GUI is based around the Spin ACD (Emboss) code and so is a bit complex to describe. If there is sufficient interest please contact me (jkbonfield) and I will post a writeup on them.

• Added new template status codes to the contig editor. E indicates that the template is consistent, but an "E"nd is missing (ie it is not a read-pair). O indicates that both ends are present, but one is in an"O"ther contig. The only "D"istance flag has now be split into "D" and "d" to distinguish between too big/small versus impossible.

• The "Sequence Search" function now searches finds matches in padded sequence when using the "by tags" mode. Matches displayed in the editor are now automatically selected to speed up tagging.

• Added an option to control the maximum alignment length displayed from find internal joins (in the text output window).

• Added some rudimentary checking for multiple gap4s writing to the same database. One gap4 should now abort (instead of corrupting the database).

• Added a new option to allow the highlight disagreements mode to be filtered by base quality.

• Ambiguity codes can now be written to FASTA files when saving the consensus.

• Added a template-size strictness option (in both the main Options menu and the template display menus) to adjust the leeway allowed in the minimum/maximum insert size information.

• Improved handling of pads in the restriction enzyme plots (sizes listed are correct for padded sequences) and stop codon plots (pads are ignored and no longer cause reading frame shifts).

• Very short sequencing vector tags may now be ignored when computing the validity of templates, as such tags are typically erroneous.

• Added the ability to filter out certain types of primer information when computing the consistency of templates. This allows workarounds for invalid PR details in the experiment files.

• Directed assembly now has the option to ignore vector-clip tags (SL/SR) in the experiment files. This means that vector tags in non-hidden data do not change the assembly when exporting and reassembling the data.

• Added a "Prev" button to the contig editors Select Primer interface.

• The editor names display can now be adjusted to show template names instead of reading names (see the editor Settings menu).

• The contig editor now remembers the last used X-scrollbar position for the names panel when restarting.

Prefinish

• Sequences containing a "FAKE" note are now ignored when computing strand coverage.

• Decreased the impact of primer3 scores on the experiment "cost".

• Added options for checking template status (-chk_template_stat) and to allow ignoring certain classes of errors.

• Added a -min_vector_len parameter so very small (and hence likely wrong) vector tags are ignored.

Other

• New command-line tool "getABIfield". This is a generic tool for dumping contents of AB1 files in a variety of formats.

• Convert_trace -normalise has been rewritten. It generally does a much better job now. It also has a "-subtract" option to remove fixed DC offsets from traces.