Version-1998.0 Release Notes

Gap4

The most significant changes in Gap4 include an interface to phrap and a corresponding consensus algorithm to use phred style confidence values, an entirely new Find Internal Joins function, and an increased limit of 32 characters in reading names (was 16).

Phred produces values on the -10log(error_rate) scale. With these confidence values Gap4 can produce the most probable consensus along with confidence values for these consensus bases. The consensus output functions can now strip the pads from the consensus sequence. This means that it is not necessary to spend time removing columns of pads when it is obvious to both users and Gap4 that there really is no base there.

The Contig Editor has also been improved to a provide considerable increase in speed of editing when dealing with the new 'confidence' consensus. The consensus can be shaded to indicate its confidence. Coupled with the "consensus quality" search it provides a means of detecting the consensus bases that are most likely to be incorrect. Gap4 can also provide an estimation of the number of errors in the consensus and an estimated error rate. This allows editing up to a desired error rate and no further. The List Confidence function gives a quick summary of the number of bases that would need to be checked in order to achieve your desired error rate.

The Find Internal Joins function has been completely rewritten. The most significant change is that it now only finds joins that reach the ends of both contigs - ie real potential joins instead of internal repeats. If you find that the old functionality of Find Internal Joins was desirable (in that it found good matches in contigs that did not extend to the ends) then we suggest the use of Find Repeats instead (which has been sped up).

Sip4

The most significant changes in Sip4 include a significant speed up in "find matching words", especially when comparing a sequence against itself, and the ability to send sequences from Gap4 and Nip4 to Sip4 and link their respective cursors together to allow synchronous scrolling between programs.

Nip4

The most significant changes in Nip4 include the additional functions: find start codons, find open reading frames, change the genetic code and count the dinucleotide frequencies.

Gap4 Changes

Main changes

• New 'confidence' consensus algorithm. This uses Phred like quality values (in the scale of -10log(error_rate)) to produce the most probable consensus base along with a consensus confidence value in the same scale. The Contig Editor has search methods for finding poor quality consensus bases. A List Confidence command (in both the main Gap4 menus and in the editor) gives estimates of the expected number of errors and frequency charts for the confidence values. Note that all the consensus algorithms now use IUB codes instead of the Staden codes.

• Phrap assembly interface to allow assembly of new sequences or reassembly of existing sequences. Requires newest version of Phrap.

• The consensus can now be output with pads removed. This includes adjusting the positions and lengths of annotations to ensure that they still mark the correct sequence fragment.

• The maximum reading name length is increased from 16 characters to 32 characters.

• New Find Internal Joins algorithm. The key feature is that it now checks that all matches are real joins rather than a possible internal repeat.

• Added a Contig Editor information line to display information on readings, the consensus, and annotations when the mouse pointer is moved over the relevant object. The contents and format of the information displayed is configurable.

• New Difference Clip and Quality Clip commands. These adjust the cutoff data based on disagreements with consensus or average quality values.

Minor updates

• The Editor Highlight Disagreements mode can now also highlight by foreground colour and can be switched into a case insensitive mode.

• The Contig Editor can now adjust individual confidence values and allows control over the default confidence values used for replaced or inserted bases.

• Improved support for compressed files in the Trace Display. Files do now not need to be renamed after compression as Gap4 will automatically search for .gz, .bz, .Z and .z extensions.

• It is now easier to assemble, disassemble or extract a single reading as this can now be done without creating a file of filenames or a list.

• Contig Editor discrepancies search now works in conjunction with auto-display traces.

• New View menu in Restriction Enzyme plot for easier access to the textual output functions.

• A sequence base confidence of 100 now forces the consensus base type to be that sequence base type. Two conflicting bases of confidence 100 will set the consensus to '-'.

• No longer do a Check Database when opening a database in read-only mode on the command line. Two new command line switches, -check and -nocheck, also control this behaviour.

• Sped up Find Repeats and Auto Assemble.

• The Join Contig command can now be used in read-only mode to view contig joins.

• Renamed the Editor "verify &" and "Verify |" searches to "Evidence for Edits (1)" and "Evidence for Edits (2)".

• Added a popup menu (containing complement, edit, etc) to the consensus ruler in the Quality, Restriction Enzyme and Stop Codon plots.

• Directed assembly now reallocs space for consensus on-the-fly hence removing the need to set maxseq before hand.

• The contig selector now displays the number of readings in a contig in its information line.

• Added a TRACE_DISPLAY.DIRECTION definition which can be either top or bottom (the default). Controls where new traces are displayed.

• Editor "Next Problem" search is now "Next Search" and can be used with any type of search.

Bugs fixed

• Ran "purify" and fixed the detected memory leaks or bad memory accesses.

• The Contig Editor sometimes would corrupt the annotation lists resulting in 'annotation neither used or free' type messages. This could result in corrupted databases.

• Fixed the "Locking an inuse record" error.

• BUSY files were incorrectly handled when opening several databases in succession without exiting the program.

• Check assembly incorrectly calculated the percentage mismatch.

• Colours on systems using CDE should now be correct.

• Better error recovery with dialogues that enter the 'busy mode' - especially for Find Internal Joins.

• Joining contigs with the multi-contig template display running was crashing.

• Search by annotation (in the Contig Editor) crashed on Solaris

• Saving tags from Find Repeats often crashed, or saved no tags.

• Over enthusiastic clicking to bring up traces was sometimes resulting in a crash.

• Occasional contig ordering crash fixed.

• Tags containing comments with very long lines were truncated when written to an Experiment File.

• The manual Join Contigs command now brings up the displays at the requested reading names rather than the contig starts.

• The old consensus algorithm dealt with dashes on the reverse strand incorrectly when in 'compare strands' mode.

• The stop codon plot now takes note of the start and end ranges in the dialogue.

• Improved the Trace Display code; better resize handling and faster.

• Editor search 'by file' mode now handles reading numbers as well as names.

• Finding names in the restriction enzyme map by moving the cursor over the cut line now works correctly.

• The restriction enzyme dialogue box would "hang" if the contig identifier was deleted and the mouse left the contig identifier entry box.

• If the quality plot or stop codon plot windows were not positioned quickly enough, the scaling of the results would go wrong.

Sip4 Changes

Main changes

• Sped up "find matching words" and especially the case of comparing a sequence against itself.

• Sequences can be sent to Sip4 from Gap4 and Nip4

• Added cursors to the sip plot and sequence display.

• Added the ability to select the sequence and sequence range to each of the comparison routine dialogue boxes. The sequence range is remembered for the next use of that sequence.

• The sip plot now contains the features of the nip plot ie the ability to move plots around, and double clicking on the raster with the middle mouse button will invoke a sequence display.

• Added the ability to create a new sequence over a specific range in the sequence manager.

Minor updates

• If a sequence is entered which has the same name as one already loaded its name is changed by the addition of '#number" where 'number' is a unique identifier.

• Previously Sip4 insisted that at least 2 sequences were loaded but now if only 1 sequence is loaded, the comparison functions will compare this sequence against itself.

• Sip4 now saves the results from "find best diagonals" and "find matching words" and treats them the same as the other comparison functions. Previously the results were only drawn to the sip plot and any changes to the plot, such as zooming, would destroy the results.

• Print out the amino acid composition if a protein sequence is entered.

Nip4 Changes

Main changes

• Added ability to select the sequence as well as the sequence range to each of function dialogue boxes. The sequence range is remembered for the next use of that sequence.

• Added a function to search for start codons.

• Added a function to search for open reading frames and save as a feature table or save the amino acids in fasta file format.

• Added a function to change the genetic code used internally within the program.

• Added a function to count the dinucleotide frequencies.

• Added ability to create a new sequence over a specific range in the sequence manager.

Minor updates

• If a sequence is entered which has the same name as one already loaded its name is changed by the addition of '#number" where 'number' is a unique identifier.

• Added a results menu to the restriction enzyme plot, and added the "Output enzyme by enzyme" and "Output ordered on position" commands.

• Start and stop codons are now plotted automatically onto an existing gene search plot.

• Added the "busy" mechanism.

Bugs fixed

• Fixed several scaling bugs which occurred when moving plots around.

• Saving to a file from the sequence manager now deals with ranges correctly.

• If all the results have been removed from a plot, it will be closed.

• String searching on the reverse strand now works correctly

Trev

Minor features

• Trace colours are now configurable in .tk_utilsrc.

• The font adjustment mechanism is now consistent with the other programs.

Bug fixes

• Shutting down using the window manager instead of the Exit command now terminates the process correctly.

• Saving an ABI or ALF file to an Experiment File was giving invalid files. Saving from SCF worked.

Others

• Added a -normalise option to makeSCF to remove trace background and to scale the peak heights.

• Extract_seq -good_only now gives the correct region.

• Improvements in reading ALF files (from any program). If the "ALF Processed data" cannot be found the "ALF Raw data" is read instead.

• Pregap now adds (if known) chemistry information during the Augment stage.

• The vector clipping program now has a fail safe (when "-m -1" is used) where the left sequencing vector cutoff is set to the distance between the primer and cut sites when no match was found. Used in pregap.

• Added ability to search for word* and *word* on entrynames in the sequence library searching dialogue.