Cnv Visualization

Question

[Bioscientist]

I've produced a list of CNV(copy number variation) data like below:

chr10    10271614    10659796    DEL
chr10    107242905    107243436    DEL
chr10    107940570    107941687    DEL
chr10    108020235    108022638    DEL
chr10    111562017    111568300    DEL
chr10    116956782    117389734    DEL
chr10    117005207    117396827    DEL

Just wondering if we have any VISUALIZATION software (say to visualize them on genome browser) so that I can check if my CNVs are overlapped? Thanks

Answer 1

[Pierre Lindenbaum]

convert your file to BED and upload it as a custom track in the UCSC genome browser.

[Larry_Parnell]

Exactly. Select a color that is different from that used by the UCSC browser for other/known CNVs.

[brentp]

+1 good idea to check against known CNVs.

[Bioscientist]

thanks a lot! cannot say enough thanks to you guys! always help me a lot

Answer 2

[brentp]

Pre-visualization, you could also use BEDTools mergeBed, something like:

mergeBed -n -i cnvs.bed > cnvs.overlap-counts.bed

where cnvs.bed is your CNV file and the final column in the output file will like the number of overlaps. So, you could get regions with more than a single CNV as:

awk '($5 > 1)' cnvs.overlap-counts.bed > overlapped.bed

Either way, IGV is also a very good tool for visualizing.

[Bioscientist]

thanks a lot! cannot say enough thanks to you guys! always help me a lot

Answer 3

[Ryan D]

Agree with Pierre above. I've pasted an example you could use below.

track name=normal description="Normal (red=deletion, green=duplication)" visibility=4 priority=1 itemRgb="On"
chr1    696961  747786  12087_N_PC     140     .       0       0       255,0,0
chr6    26845010        26876548        12087_N_PC     260     .       0       0       128,0,0
track name=tumor description="Tumor (red=deletion, green=duplication)" visibility=4 priority=1 itemRgb="On"
chr18   67191847        67206726        12087_T_PC      0       .       67191847        67206726        0,128,0
chr20   54595851        55224865        12087_T_PC      0       .       54595851        55224865        0,128,0

Answer 4

[Sehi L'Yi]

If the file is not too large, you could use the file as-is, store them somewhere, and use its URL on http://gosling.js.org.