Need fore speed...

[Beynon, Rob]

I don't want to be accused of passing effort out to the list, but I would like to know, from the experts, the fastest way to execute some code..

 

I have a file of tab delimited items like so...16,000 lines long read into a single variable. (The right bracket is part of the file, not a quote mark)

 

>DAPA_ECO57   MFTGSIVAIVTPMDEKGNVCR      2267.1049     1

>DAPA_ECOL6   MFTGSIVAIVTPMDEKGNVCR      2267.1049     1

>DAPA_ECOLI   MFTGSIVAIVTPMDEKGNVCR      2267.1049     1

>(-M)YEGS_ECOLI      AEFPASLLILNGKSTDNLPLR      2268.2398     1

>GLTS_ECOLI   MFHLDTLATLVAATLTLLLGR      2269.28372    0

>(-M)YDIV_ECOLI      KIFLENLYHSDCYFLPIR   2270.1462     1

>YBAA_ECO57   MKYVDGFVVAVPADKKDAYR 2271.16256    3

>YBAA_ECOL6   MKYVDGFVVAVPADKKDAYR 2271.16256    3

>YBAA_ECOLI   MKYVDGFVVAVPADKKDAYR 2271.16256    3

 

I now want to search each line, and match an input number (obsMass) against the third item of each line (pepMass)

If it matches, within a certain tolerance (expressed in parts per million (ppm)  - the tolerance changes with the magnitude of the number) I want to do something with that line. Ties (or multiple hits to the same obsMass) are OK.

 

I then want to repeat that process for several thousand different values of obsMass. In other words, I will evaluate the mass matching loop tens of millions of times.

 

Some questions - is there advantage in sorting the list by mass first?

Would I gain speed by using integer arithmetic (is that even possible?) or by matching the numbers as strings?

 

And most importantly..

Is there an elegant rev way of handling this that I don't even know about?!

Thanks

Rob

 

 

________________________________

Prof R J Beynon[h]
Proteomics and Functional Genomics Group
Faculty of Veterinary Science
University of Liverpool
Crown Street, Liverpool L69 7ZJ

________________________________

Phone: +44 151 794 4312

Fax: +44 151 794 4243

Email: r.beynon at liv.ac.uk

http://www.liv.ac.uk/pfg

________________________________