| 1 | #!/usr/bin/python2.6 |
|---|
| 2 | |
|---|
| 3 | """ |
|---|
| 4 | Reads a list of intervals and a maf. Produces a new maf containing the |
|---|
| 5 | portions of the original that overlapped the intervals |
|---|
| 6 | |
|---|
| 7 | NOTE: See maf_extract_ranges_indexed.py which works better / faster for many |
|---|
| 8 | use cases. |
|---|
| 9 | |
|---|
| 10 | TODO: Combine with maf_extract_ranges, and possibly share some code with |
|---|
| 11 | maf_extract_ranges_indexed. |
|---|
| 12 | |
|---|
| 13 | usage: %prog interval_file refname|refindex [options] < maf_file |
|---|
| 14 | -m, --mincols=10: Minimum length (columns) required for alignment to be output |
|---|
| 15 | -p, --prefix=PREFIX: Prefix |
|---|
| 16 | """ |
|---|
| 17 | |
|---|
| 18 | import psyco_full |
|---|
| 19 | |
|---|
| 20 | from bx.cookbook import doc_optparse |
|---|
| 21 | |
|---|
| 22 | import bx.align.maf |
|---|
| 23 | from bx import intervals |
|---|
| 24 | import sys |
|---|
| 25 | |
|---|
| 26 | |
|---|
| 27 | def __main__(): |
|---|
| 28 | |
|---|
| 29 | # Parse Command Line |
|---|
| 30 | |
|---|
| 31 | options, args = doc_optparse.parse( __doc__ ) |
|---|
| 32 | |
|---|
| 33 | try: |
|---|
| 34 | range_filename = args[ 0 ] |
|---|
| 35 | try: |
|---|
| 36 | refindex = int( args[ 1 ] ) |
|---|
| 37 | refname = None |
|---|
| 38 | except: |
|---|
| 39 | refindex = None |
|---|
| 40 | refname = args[ 1 ] |
|---|
| 41 | if options.mincols: mincols = int( options.mincols ) |
|---|
| 42 | else: mincols = 10 |
|---|
| 43 | if options.prefix: prefix = options.prefix |
|---|
| 44 | else: prefix = "" |
|---|
| 45 | except: |
|---|
| 46 | doc_optparse.exit() |
|---|
| 47 | |
|---|
| 48 | # Load Intervals |
|---|
| 49 | |
|---|
| 50 | intersecters = dict() |
|---|
| 51 | for line in file( range_filename ): |
|---|
| 52 | fields = line.split() |
|---|
| 53 | src = prefix + fields[0] |
|---|
| 54 | if not src in intersecters: intersecters[src] = intervals.Intersecter() |
|---|
| 55 | intersecters[src].add_interval( intervals.Interval( int( fields[1] ), int( fields[2] ) ) ) |
|---|
| 56 | |
|---|
| 57 | # Start MAF on stdout |
|---|
| 58 | |
|---|
| 59 | out = bx.align.maf.Writer( sys.stdout ) |
|---|
| 60 | |
|---|
| 61 | # Iterate over input MAF |
|---|
| 62 | |
|---|
| 63 | for maf in bx.align.maf.Reader( sys.stdin ): |
|---|
| 64 | if refname: |
|---|
| 65 | sourcenames = [ cmp.src.split('.')[0] for cmp in maf.components ] |
|---|
| 66 | try: refindex = sourcenames.index( refname ) |
|---|
| 67 | except: |
|---|
| 68 | continue |
|---|
| 69 | |
|---|
| 70 | ref_component = maf.components[ refindex ] |
|---|
| 71 | # Find overlap with reference component |
|---|
| 72 | if not ( ref_component.src in intersecters ): continue |
|---|
| 73 | intersections = intersecters[ ref_component.src ].find( ref_component.start, ref_component.end ) |
|---|
| 74 | # Keep output maf ordered |
|---|
| 75 | intersections.sort() |
|---|
| 76 | # Write each intersecting block |
|---|
| 77 | for interval in intersections: |
|---|
| 78 | start = max( interval.start, ref_component.start ) |
|---|
| 79 | end = min( interval.end, ref_component.end ) |
|---|
| 80 | sliced = maf.slice_by_component( refindex, start, end ) |
|---|
| 81 | good = True |
|---|
| 82 | for c in sliced.components: |
|---|
| 83 | if c.size < 1: |
|---|
| 84 | good = False |
|---|
| 85 | if good and sliced.text_size > mincols: out.write( sliced ) |
|---|
| 86 | |
|---|
| 87 | # Close output MAF |
|---|
| 88 | |
|---|
| 89 | out.close() |
|---|
| 90 | |
|---|
| 91 | if __name__ == "__main__": __main__() |
|---|