first previous next last contents

Checking for Errors

It is important to check the final sequence for any errors introduced by incorrect editing. We strongly advise this when making use of the more dangerous options in "Edit Modes" as it is possible to accidentally make changes. The editor provides several methods for checking the edits performed on the data.

The search menu contains two search types listed as "Evidence for edits (1)" and "Evidence for edits (2)". "Evidence for edits (1)" searches for the next edited place where none of the original sequences (as from the base caller) at this point agree with the consensus. This helps to spot cases where entire columns have been inserted or deleted. "Evidence for edits (2)" performs the same checks as "Evidence for edits (1)", but on each strand independently. So it will find all edited places where none of the original sequences in a strand agrees with the consensus for that strand.

Further checks may be performed outside the editor. Using the Find Read Pairs command all templates containing both forward and reverse readings will be checked to make sure that the relative orientation and distance of the sequences is correct. See section Find Read Pairs.

Finally, the Check Assembly command can either check the hidden data for each reading to check that it does not diverge from the consensus sequence, or the visible data can be examined to locate segments with a high proportion of disagreements with the consensus. See section Check Assembly. Such cases arise from unnoticed section of vector sequence, chimeric reads, or due to a reading being in the wrong copy of a repeated element.


first previous next last contents
This page is maintained by James Bonfield. Last generated on 2 Febuary 1999.
URL: http://www.mrc-lmb.cam.ac.uk/pubseq/manual/gap4_72.html