Other random tips

Extract text from PDF

Tools Tried

Following are based on my attempts to extract phenotype data from phenotips file. These may not necessarily be true/apply for other pdf files.

pypdf2

Python module. Failed to extract most of the text. Extracted just small part of pdf.

pdfminder

Command line tool
Required installation from source.
Worked reasonably well. Seemed to extract most of the text, if not all.
Has -layout_mode option to preserve the layout of text as in pdf. Haven't tried though.
Doesn't seem to allow ignoring header/footer though

pdftotext

Command line tool
Installed in mac - "brew install poppler". It's part of poppler suite.

# simple conversion
pdftotext sample.pdf out.txt

# to keep layout of text as in input pdf
pdftotext -layout sample.pdf out.txt

# to change encoding. Default UTF-8 encoding resulted in 'fi' as one letter instead of twletters. ASCII7 worked out well.
pdftotext -layout enc ASCII7 sample.pdf out.txt

Recognizing header and footer is not straightforward but can be done by choosing the area of pdf tthat need to be converted. See this stackoverflow answer.
What worked for me in my case for phenotips file

pdftotext -layout enc ASCII7 -y 80 -H 640 -W 1000 sample.pdf out.txt

Convert Excel xlsx to csv

xlsx2csv
Easy to install; Worked well in my try.
More tools

UCSC genome browser

Interesting features

See settings used in text format

Either of these would do:

Replace existing url with http://genome.ucsc.edu/cgi-bin/cartDump to see the settings used in text format
Save settings file via my data -> my session -> save settings

Short match

BLAT is useful is seq is >20 bases. For shorted sequences (2-30 bases), short match should work better.

Show DNA seqs in color based on features

Can be used to color and toggle case for sections of DNA (exon, SNP, etc.)

Primer design

Save snapshots

View -> PDF/PS
Change hgTracks to hgRenderTracks in url to get the png output http://genome.ucsc.edu/cgi-bin/hgRenderTracks?parameters