genome browser and bigwig file issue!

[emolinari]

Hi everyone,

it is the first time that I try to visualize my data on the UCSC Genome Browser.
I am not so familiar with the data format required for custom track visualization so I have followed some indications from papers, manuals etc...

I have 20 RNAseq samples (2x75, ~60 mil reads each, Homo Sapiens).

here's the pipeline I have followed:

bedtools bamtobed -split -i accepted_hits.bam > accepted_hits.bed
sort -k 1,1 accepted_hits.bed > sorted.accepted_hits.bed
bedtools genomecov -i sorted.accepted_hits.bed -bg -scale 10 -g chrom.sizes > normalized.bg ###so that I could normalize for 10 RPM###
bedGraphToBigWig normalized.bg chrom.sizes mybigwig.bw

The file are uploaded on the genome browser and I don't see any error message BUT I don't see anything at all...zero!
I am almost sure I have done something very stupid, maybe in creating the files or during the visualization process.

Can anyone help???

Thanks!
Manu

[GenoMax]

Have you looked at the file contents to assure yourself that the results at each step look reasonable.

Custom tracks display is not set to "hide", correct? Navigate to a gene you know should be there and check the contents in that region.

[emolinari]

Have you looked at the file contents to assure yourself that the results at each step look reasonable.

Custom tracks display is not set to "hide", correct? Navigate to a gene you know should be there and check the contents in that region.

Hi GenoMax

I did put it on dense and yes, no sign of any gene -not even housekeeping.
I've been looking at my files like crazy...I have a feeling that it is related to the chromosome name -still weird I don't see any error message.

I am trying to fix it with hg19 fetchchromsizes...

let's see what I get!

Manu

[emolinari]

Hi GenoMax

I did put it on dense and yes, no sign of any gene -not even housekeeping.
I've been looking at my files like crazy...I have a feeling that it is related to the chromosome name -still weird I don't see any error message.

I am trying to fix it with hg19 fetchchromsizes...

let's see what I get!

Manu

...ta-da!

yep, the chromosome name was indeed wrong...and yep, there was an error message that I couldn't see!

hop this could be helpful for any newbie as me!!!

Manu

[Udiland]

chromosome sizes for arabidopsis

Hi,

I am trying to make a bedgraph file form a ChIP-Seq bam file using genomeCoverageBed.

I don't know what to put in the genome argument of the fetchChromSizes command in order to get the chromosome sizes of Arabidopsis Thaliana

Thank`s

[GenoMax]

Hi,

I am trying to make a bedgraph file form a ChIP-Seq bam file using genomeCoverageBed.

I don't know what to put in the genome argument of the fetchChromSizes command in order to get the chromosome sizes of Arabidopsis Thaliana

Thank`s

Since UCSC does not host a copy of the Arabidopsis genome where are you trying to get the sizes from?

You could make up a simple file like this (example for A. thaliana 167 assembly).

chr1 	30427671  
chr2 	19698289  
chr3 	23459830  
chr4 	18585056  
chr5 	26975502  
chrC 	154478  
chrM 	366924

[Udiland]

Thank you. I understand that it is not advisable to make my own chromosome size file. ( I think it is because the format needed is more complicated than what you wrote).

[GenoMax]

*.chrom.sizes file is not complicated (see this example for hg19 from UCSC: http://genome.ucsc.edu/goldenpath/help/hg19.chrom.sizes).

You would want to make sure that your file matches reference genome build you are using but that should be it.

Note: You can get the information you need from SAM file headers (if you have already done the alignment).

[Udiland]

Hi,

Thank you. somehow it didn't work.
when I use a file with the header (i.e "chrom""size") I get this message:
invalid unsigned integer: "size"

and when i don't use header (like in the humen example you referred to) I got this message:
1 is not found in chromosome sizes file

[GenoMax]

Are your chromosome names matching (e.g. chr1 as opposed to 1)? Is there a tab separating the column values?

Can you post the header lines from your aligned SAM/BAM file?

[Udiland]

This is my *.chrom.sizes file (it is tab separated)
chrom size
chr1 30427671
chr2 19698289
chr3 23459830
chr4 18585056
chr5 26975502
chrC 154478
chrM 366924

and this is my sam file header
@HD VN:1.0 SO:unsorted
@SQ SN:1 LN:30427671
@SQ SN:2 LN:19698289
@SQ SN:3 LN:23459830
@SQ SN:4 LN:18585056
@SQ SN:5 LN:26975502
@SQ SN:Mt LN:366924
@SQ SN:Pt LN:154478
@PG ID:bowtie2 PN:bowtie2 VN:2.0.6

[GenoMax]

Remove "chr" from names in your chrom.sizes file. As you can see in your sam file your reference chromosomes are named 1,2,3 (and not chr1, chr2).

chrM becomes Mt. Add an entry for Pt and remove ChrC.

Let us know if that works.

[Udiland]

it worked!!

but without the header..

thanks a lot.