Wierd SAM format chromosome column

[serenaliao]

Dear all,

According to my previous experience, in sam file output by bismark, the third column is usually chromosome number with "chr" in front. While my new data has a mixture of chr<number> and <number> in that column. Is that common?

MWR-PRG-0014:106:C25B0ACXX:4:1101:1195:1983_1:N:0:CAAAAG/1 115 15 3399663 255 100M = 3399567 -196 CAAAAT
ACCAAAAAATAAAACACAAAATAAAAAAACTATTCTTCCTACCTAAAAACATAATAACTTCCACATCAATAATTCTTTATTACATAAATTATAN #DDDC@</=BCEEEC>ECCD@B@?7EA=HIIIIGBIHG
F>ACFFF??GGIIIHGCDD<<<B?19CIGHHIIIEHHEIHEIGIGEDEHHFHDHDDDBB=1# NM:i:23 XX:Z:2G3G3G1GGG1G1G5GG4G2GG12T2G1G1G1C2G1G31G4GT XM:Z:.
.x...h...x.hhh.h.h.....xh....h..hh...............x.h.h....h.h...............................h....h. XR:Z:CT XG:Z:GA
MWR-PRG-0014:106:C25B0ACXX:4:1101:1195:1983_1:N:0:CAAAAG/2 179 15 3399567 255 100M = 3399663 196 AACAAA
AACAATTATAAAACTAAACTAAAAAAATCCCAAATCAAAATTTTAATATTAATTTATTCATTCACCTCACAAATAAATAAAAATATTTATCAAA B@@DD;DDFDDHDEBFHIIGGIJJIJJIJIIE>BD<?F
DFGBGGID@CGICCGGIJJJJIJJJI=AECHHAE>BFFCCEE@ECCDAAC@=C>CCCDC;AC NM:i:24 XX:Z:1G2GGG3G2G1GGG3GG3GG1G8GG4G5G2G2G25G4G6G3G1 XM:Z:.
h..xhh...x..h.hhh...xh...xh.h........xh....h.....h..h..h.........................h....h......h...x. XR:Z:GA XG:Z:GA
MWR-PRG-0014:106:C25B0ACXX:4:1101:1234:1999_1:N:0:CAAAAG/1 115 chr9 57387259 255 100M = 57387164 --195 AAACAAGCATCTTAAAATAACTATTAAAATTCAAAAAACTATATATCCTCAAAACTAAAAATAATATCAAATCCATAATCTTAAAATCCTCTTTCTAAGN ?A>5>;-.;(.6:EEDA>?@
?=;DDDDACIIDECB=<??DBBEDDDDDD4DBIDIFCDD>EE?C9BDB<4<DEFEFEAEBDC+A9<B>DDBDB=A11# NM:i:21 XX:Z:1G3G8GG2GG2G2G7GG1G14G1G3G2G1G2G4G11GG15T
XM:Z:.h...xH.......hh..hh..x..h.......xh.h..............x.h...h..h.h..h....h...........hh..............H. XR:Z:CT XG:Z:G
A
MWR-PRG-0014:106:C25B0ACXX:4:1101:1234:1999_1:N:0:CAAAAG/2 179 chr9 57387164 255 100M = 57387259 -195 CAAAATATTTATACTACCTACTATATACACAACACAATACTAAATTCCACTTTACTCCCTAATCTTCCACTATTCCCTCTTCCCTAAACAAAAAAAAACA ?@@FF?D4=E?FF3:AA:
C4C:FGG>3+AFHIJGDH@DHGBDHIGG?@<FGGGHJIDCDHCHBGGIIJIIBEE=EHE;A?E7;7?7;;AA@BDD@? NM:i:19 XX:Z:6G3G1G2G3G2G1G1G4G3GG1G3G10G7G25G4G1G1G3-XM:Z:......h...h.h..x...x..x.h.h....x...zx.h...h..........h.......h.........................h....h.h.h... XR:Z:GA XG:Z:GA
MWR-PRG-0014:106:C25B0ACXX:4:1101:1406:1986_1:N:0:CAAAAG/1 115 12 77985032 255 100M = 77984894 --238 AATATTAACATAAAACCAAAACAACAAAATATCTAAAACACTCCAATCCCCACTCATTCCAAACTTCAAACTACTAAATCAAAAATATTACATTTCATTN CDAEDEED@C>DDBB@EFFFD@
HHCHHGGECCAEHD=.=8HFJHIHCHEGB9IIIHGGGIJIGIGDGIHHGEEJEJGIGIJJJJJIJHHFGHFFFDD=4# NM:i:30 XX:Z:3G2GG1G1GG1G3GG2GG3GG1G3GG8G16G6GG5GGG3GG
GG1G2G9A XM:Z:...h..hh.z.hh.h...xh..zx...hh.h...xh........z................x......xh.....xhh...xhhh.h..h.......... XR:Z:C
T XG:Z:GA

[dpryan]

That's odd, you might run the following on your reference fasta file to see if this is expected or not:

grep ">" reference_genome.fa

If ">15" pops up, then this is normal, though it'd be odd to have that and chr9 in the same fasta file. bismark does play around a bit with contig names, but something being messed up in the code dealing with that should result in different behaviour.

[fkrueger]

That's odd, you might run the following on your reference fasta file to see if this is expected or not:

grep ">" reference_genome.fa

If ">15" pops up, then this is normal, though it'd be odd to have that and chr9 in the same fasta file. bismark does play around a bit with contig names, but something being messed up in the code dealing with that should result in different behaviour.

Bismark takes whatever the fasta files had in the header until it hits the first white space, if you get '15' and 'chr9' in the output I would assume that these entries looked like '>15' and '>chr9' in the fasta files you used for the genome indexing process. I think it does replace '|' characters with underscores, but it would certainly not add or remove 'chr'.

[serenaliao]

Bismark takes whatever the fasta files had in the header until it hits the first white space, if you get '15' and 'chr9' in the output I would assume that these entries looked like '>15' and '>chr9' in the fasta files you used for the genome indexing process. I think it does replace '|' characters with underscores, but it would certainly not add or remove 'chr'.

Thanks fkrueger,

You are right. This happened to my FASTA file.(some are chr<number> and some are <number>) Is there any convenient way to add "chr" before the chromosome number in SAM file(third column) if there is no chr? Thanks!

[serenaliao]

Thanks fkrueger,

You are right. This happened to my FASTA file.(some are chr<number> and some are <number>) Is there any convenient way to add "chr" before the chromosome number in SAM file(third column) if there is no chr? Thanks!

Just to follow up, I used awk '{if($3!~/^chr/){$3="chr"$3} print($0)}' filename. Does this sound reasonable?

[fkrueger]

I am no expert with awk but it looks ok, should be easy enough to test (maybe on a few lines first). Any clues why your fasta files have mixed chromosome names?