Rare clusters have all "n"s in 4th field, causing R to fail

[tedsharpe]

Of the 359,984 clusters produced by processing the 1000 genomes cram for sample HG00150, 5 have nothing but "n" calls in the 4th field. Here are two examples:
chr1:24599821 left 7 nnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnn ccccccccccccccccccttcgttaacgacgctattaccaactaaa
chr2:232427156 left 5 nnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnn ttttttttttttttttttttttttttttttttttttctctccccccccccccccccctctccctttcttgtgtttttttttttttttttttttccgcctcccccccacctcccaggtt

When the R script attempts find deletions using such a cluster, the analysis fails with this error:
Error in width(strings) : NAs in 'x' are not supported
Calls: do.dels ... .charToXStringSet -> solveUserSEW -> width -> width

My workaround is to clean the cluster file of these few bad guys before processing, but perhaps do.dels.R could exclude them.

Thanks for the software.