First of all, where are you data come from? It seems inconsistent! Why the coordinates are noted as strings?

NM_198212 chr7 + 115926679 115935830 115927071 11593344 2 115926679,'115933260', 115927221,'115935830',

Nevertheless, in python you can split every string (as suggested by Zev) and the get your data of interest from the list

for line in open( fname ):
    dataList = line.split('\t')
    name,chrom,strand,start,stop = dataList[:5]
    print name,chrom,strand,start,stop

If you have GenBank file as input, I recommend Biopython!

Try to do it in ipython - you can test your code:)

I got an headache just by looking at your regular expression ;) .

Seriously: You are over complicating. I suggest the following:

1. Remove commas and apostrophes. Use, say str.replace(",", "")
2. Split on space (or eventually tab) to get the tokens toks = str.split(" ")
3. On toks[2] you have the strand, toks[3] and [4] have start and end
4. toks[7] gives you the number of exons, and from there you can extract initial and starting positions and go on with strand logic.

Seasoned programmers struggle with writing (and reading) reg exps like that. Furthermore you would have to make it more complex than it is currently: you only support two exons as it stands, you need to repeat the whole group (+ operator).

Replace the reg ex with split logic. I recommend to make your like less complicated

OK: suppose this is chromsome 25:

AAAAAAAAAACCCCCCCCCCTTTTTTTTTTGGGGGGGGGG

(there are 10 of each character).

Now: if I am looking for an unspliced gene on: <blah> chr25 + 10 20 <blah>

Then the gene starts on position 10 (starting from 0), and goes up to but not including position 20. So its:

CCCCCCCCCC

This is easy. It matches python string slicing really well.

Its more confusing if I give you:

<blah> chr25 - 10 20 <blah>

What you have is the positive strand. But this gene is on the negative (complementary) strand. Remember what the chromosome looks like as a souble-strand:

AAAAAAAAAACCCCCCCCCCTTTTTTTTTTGGGGGGGGGG TTTTTTTTTTGGGGGGGGGGAAAAAAAAAACCCCCCCCCC

We are looking for the gene on the bottom strand. Meaning we count from 0 starting on the right. Number the top strand from the left, and the bottom strand from the right. So what I want here is AAAAAAAAAA.

The catch is that I'm only giving you the top strand. I'm not giving you the bottom strand. (You could generate yourself from the top strand — but given how large it is, I advise against that.)

So you need to convert coordinates. On the bottom strand, base 0 (the right-most C) is opposed to base 39 on the top strand. Base 1 is against base 38. Base 2 is against case 37. (Important point: notice what happens when you add these two numbers up — every time.) So base 10 is against base 29, and base 19 is against base 20.

So: if I want to find base 10-20 on the bottom strand, I can look at base 20-29 on the top (and then reverse-complement it).

I need to figure out how to translate to coordinates on the bottom strand to the equivalent coordinates on the bottom strand. Yes: it is very confusing

I have tried:

fields = line.split(' \t')
geneID, chr, strand = fields[:2]
start = int(fields[3])
end = int(fields[4])
if strand == '-':
     newstart,newend = -(start +1), -(end +1)

Now I have to combine this with the function below to get the reverse compliment, right?

which is on the right track, yet but its not enough. This would take the 10 and 20, and turn it into a 20 and 10.

And I know I can reverse complement the string by doing this:

r = s[::-1]
bc = {'A': 'T', 'C': 'G', 'G': 'C', 'T': 'A'}
l = list(r)
o = [bc[base] for base in l]
     return ''.join(o)

But, from here I am lost.