update bed parsing

slightly faster IntervalTree.from_tuples than .addi

README.md

@@ -52,6 +52,7 @@ Use Truvari's comparison engine to consolidate redundant variants in a merged mu
  - [segment](https://github.com/acenglish/truvari/wiki/segment) - Normalization of SVs into disjointed genomic regions
  - [stratify](https://github.com/acenglish/truvari/wiki/stratify) - Count variants per-region in vcf
  - [divide](https://github.com/ACEnglish/truvari/wiki/divide) - Divide a VCF into independent shards
+ - [ga4gh](https://github.com/ACEnglish/truvari/wiki/ga4gh) - Consolidate benchmarking result VCFs
 
 ## 🔎 More Information
 

truvari/annotations/density.py

@@ -61,8 +61,8 @@ def density_main(args):
     # setting new indexes after masking
     new_tree = defaultdict(IntervalTree)
     cnt = 0
-    for chrom in tree:
-        for intv in tree[chrom]:
+    for chrom, intvs in tree.items():
+        for intv in intvs:
             for i in range(intv.begin, intv.end, args.stepsize):
                 new_tree[chrom].addi(
                     i, min(intv.end, i + args.windowsize), data=cnt)

truvari/region_vcf_iter.py

@@ -103,7 +103,7 @@ def build_anno_tree(filename, chrom_col=0, start_col=1, end_col=2, one_based=Fal
     """
     idx = 0
     correction = 1 if one_based else 0
-    tree = defaultdict(IntervalTree)
+    ttree = defaultdict(list)
     for line in truvari.opt_gz_open(filename):
         if line.startswith(comment):
             continue
@@ -115,8 +115,11 @@ def build_anno_tree(filename, chrom_col=0, start_col=1, end_col=2, one_based=Fal
             m_idx = idxfmt.format(idx)
         else:
             m_idx = idx
-        tree[chrom].addi(start, end + 1, data=m_idx)
+        ttree[chrom].append((start, end + 1, m_idx))
         idx += 1
+    tree = {}
+    for chrom in ttree:
+        tree[chrom] = IntervalTree.from_tuples(ttree[chrom])
     return tree, idx