I would like to calculate the mappability of 50 kb windows of Mus musculus assembly 9 (mm9). I downloaded the mappability with 100 bp reads from UCSC and created windows with bedtools makewindows -g myGenomeFile.txt -w 50000 > windows.bed

I cannot just calculate the mean mappability of all overlapping elements in the mappability track for each feature in the windows.bed file because the mappability windows from UCSC have different lenghts. I believe the UCSC binary bigWigAverageOverBed is specifically made for this job. However, I keep getting segmentation faults with this binary when I try it. I cannot find any solutions for this segmentation fault. What are my alternatives/options?

Many thanks in advance