tophat alignment score and parameters

[aevgup]

I am using tophat2 for my rna seq data set. I mostly always use default options for tophat. I have some doubts when i see the accepted.hits.bam file

In the tophat manual it is written that

--max-insertion-length <int> The maximum insertion length. The default is 3


--max-deletion-length <int> The maximum deletion length. The default is 3

Now this means that in the CIGAR string in the alignment file i should never get a CIGAR string which has something like 4D or 5I because the default is 3 and these are max insertion and deletion length

But i still get something like this 39M5I4M1D28M or 71M4D5M

Am i missing something??

Secondly we have AS:i in tophat bam files. What is the best alignment score??? When i see bam files i see scores like AS:i:-29 and also like AS:i:-6 , AS:i:0

PS: isubmit tophat runs with my own genome created for 80 genes i am interested in. Can this cause the above problem??. I don't supply a gtf/gff file??

Hope to hear soon

Regards

Varun