Edit distance sequence tags available (link within)

[bcf]

Hi everyone,

I've generated several sets of edit (levenshtein) distance/metric sequence tags from 4nt to 10nt and edit distance 3 to edit distance 9. Within each sheet of the spreadsheet (or length/edit distance combination in csv format), all tags are at least the given edit distance from one other.

While not entirely exhaustive (for several reasons), these sequence tags approach the maximum number possible (after filtering) within each category of length.

I'm in the process of prepping a manuscript describing the code for tag generation (as part of a larger package on which i'm working - largely in a private branch of the repo.), but it seemed these tags might be useful to others in the meantime.

the tags are available from:

https://github.com/BadDNA/edittag/downloads

there's a blog post with a bit of description here:

http://b.atcg.us/blog/2010/11/20/a-plethora-of-sequence-tags.html

and the code to generate the tags is here:

https://github.com/BadDNA/edittag/blob/master/edittag/make_levenshtein_tags.py

I generated the tags in the files provided using the following (to generate 10nt tags with a minimum edit distance of 3, for example):

python make_levenshtein_tags.py --tag-length=10 --edit-distance=3 \
--no-polybase --gc --comp --use-c --multiprocessing \
--min-and-greater | tee 10_nt.txt

Run time for this example was 80 hours or so on 6 cores of 8 (Mac Pro, 8 GB RAM) or approx. 480 hours wall-clock time. Run times for smaller tag sets (e.g. 8 nt) are about a minute.

if you use these tags, for the time being, please "cite" the repository link (https://github.com/baddna/edittag/). Once we get a manuscript out, I'll provide proper citation information in the README.

WORDS OF CAUTION: i'm undertaking a re-organization of the code in this repository, committing lots of unit tests, refactoring, documenting, etc. - so expect the code to change as i update and refactor.

please let me know if you have any questions.

best,
b

[bcf]

edittag packaged up and on pypi

Hi everyone,

two quick things:

• edittag is now up on pypi: http://pypi.python.org/pypi/edittag
• a preprint is available from: http://precedings.nature.com/documents/5672/version/1

-b