Cigar String really high N

Question

[nicolas.ares]

I'm trying to find gene fusions, and when looking at the reads aligned by star I find that some reads will map 3 bases on a gene and then have no other base correct. When I look at the cigar I see the following 3M<high number > 400000bp>N47M. To see this reads I have used samtools view, why does this happen. This can be an example:

Analising fusion:  RP11-96H19.1::RP11-446N19.1

READS IN BAM

['SRR064438.11916954', 83, 12, 46387924, 255, '49M1S', '=', 46387805, -168, 'TAAGACCAGACCAAATCAAACCAAACCAAGCAAACCACGGGGAATGGAGA', 'A@@>9BBA@@B@B@B@BAB>B??@@BCCBABBB@BB@B?BBABBCBACAB', 'NH:i:1', 'HI:i:1', 'AS:i:97', 'nM:i:0', 'NM:i:0']

['SRR064439.375905', 163, 12, 46387934, 255, '39M11S', '=', 46652514, 264630, 'CCAAATCAAACCAAACCAAGCAAACCACGGGGAATGGAGATTATTGCCTG', 'ABBBBB@BBBAABBB@@BBBABBB@@BABBBB@ABBBACBBBBBBB=5BB', 'NH:i:1', 'HI:i:1', 'AS:i:84', 'nM:i:0', 'NM:i:0']
***
**['SRR064438.12267353', 99, 12, 46387943, 255, '30M264417N20M', '=', 46652487, 264594, 'ACCAAACCAAGCAAACCACGGGGAATGGAGATTATTGCCTGCTCCTCCAA', 'BBBBBBA@BBB?AAB@?B?A@@A@AAAA<A?@A@A?<<8=*<@=5><8<=', 'NH:i:1', 'HI:i:1', 'AS:i:95', 'nM:i:0', 'NM:i:0']
['SRR064439.1473779', 163, 12, 46387962, 255, '11M82180N39M', '=', 46874958, 487046, 'GGGGAATGGAGGTCATGTGAGCACACAGCATAAAGGCAGCTGCCCACAAG', 'BCBBCCCBBCB@:CBBB>>BBACABCBBC?ACBC<ABBA94?:7>AA9B;', 'NH:i:1', 'HI:i:1', 'AS:i:97', 'nM:i:0', 'NM:i:0']
['SRR064439.2800317', 355, 12, 46387970, 3, '3M486844N47M', '=', 46874828, 486908, 'GAGGACCTGATGATTGATTTAGCATCTTTGGCATCCGGCCACTGCTCTGC', 'B@BAAABB>?A@@AAA?BBAB>A?@@BB???BA@B@::@AAA<<B=@;:;', 'NH:i:2', 'HI:i:2', 'AS:i:97', 'nM:i:0', 'NM:i:0']**
***
['SRR064438.4051038', 99, 12, 46652386, 255, '7S43M', '=', 46652461, 123, 'GGGGACTACAGATTATTGCCTGCTCCTCCAAGCCCTTCACTGTAGAATGG', 'BBBB@=BAAAB@BB@BBB<;?A?B?8@@=@@@?8;B@<;<BB??B;?@A@', 'NH:i:1', 'HI:i:1', 'AS:i:89', 'nM:i:0', 'NM:i:0']


REFERENCE:

TAAGACCAGACCAAATCAAACCAAACCAAGCAAACCACGGGGAATGGA==GGTAGGTGAATAGCGCCAAAGAGAATGATGGCTCACAACACTTCTAAGCA

READS:

TAAGACCAGACCAAATCAAACCAAACCAAGCAAACCACGGGGAATGGA==GA
***
                    CCAAATCAAACCAAACCAAGCAAACCACGGGGAATGGA==GATTATTGCCTG

                             ACCAAACCAAGCAAACCACGGGGAATGGA==GATTATTGCCTGCTCCTCCAA

                                                GGGGAATGGA==GGTCATGTGAGCACACAGCATAAAGGCAGCTGCCCACAAG
***
                                                        GA==GGACCTGATGATTGATTTAGCATCTTTGGCATCCGGCCACTGCTCT GC

The 2,3,4th reads are the ones that bring up this problem, should I account them as mapping here or not? Is there something I'm missing?

Answer 1

[rpolicastro]

When aligning RNA-seq with STAR a cigar string like 30M264417N20M usually means: 30 mapped bases (exonic), followed by a 264417 bp intron, and then another 20 mapped bases (exonic).

This was likely a 50bp illumina sequencing read that spanned an exon-exon junction, so when you align it to the genome there will be a huge gap where the intron was.