Circos:- How To Make Circos Histogram Image Using This Data

Question

[ajaybioinfo]

Dear all,

i am very much new to Circos, able to generate circos image of any table data type but now my supervisor is asking of generating the genome wide circos image of the R and DR genes distributed on the 11 rice chromosome (position specific) . i try to learn from the tutorial but could not able to understand how draw the image, what i have learn - for the table data image generation that we need the data in the table format .csv or separated by space and the file is saved in the .txt format after that from the circos tool we have to use parse-table and finally make-conf after that final .conf file generation followed by the circos -conf command which generate the image file.

please tell me what are the steps i have to follow and which tool i have to use to prepare the conf file. i am very much thankful to your for your kind guidance and support for your reference please find the below data format which i have to plot using circos

Chromosome     Start    End    Function    
Chr01    176373    175717    NBS-LRR    
Chr01    228127    228513    Defence    
Chr01    263322    262761    LRR-TM    
Chr01    343320    345126    NBS-LRR    
Chr01    476287    476105    LZ-NBS-LRR    
Chr01    689688    688555    LRR    
Chr02    769100    768793    NBS-LRR
Chr02    787353    786342    NBS-LRR
Chr02    809973    811358    NBS-LRR
Chr02    822559    824064    NBS-LRR
Chr03    833569    832507    NBS-LRR
Chr03    849028    850740    LRR-TM
Chr03    878123    878650    LRR-TM
Chr03    886993    888963    LRR-TM
Chr03    892021    890924    NBS-LRR
Chr04    895807    896879    NBS-LRR
Chr04    883367    882194    LRR
Chr05    903850    903126    NBS-LRR
Chr05    916608    915502    NBS-LRR
Chr05    920942    922017    LRR-TM
Chr06    13095571    13098112    LRR
Chr06    13281335    13280944    LRR
Chr07    13326803    13328111    LRR-TM
Chr07    13803509    13804285    LRR
Chr08    16807025    16806060    Defence
Chr08    16816957    16816029    Defence
Chr08    16820145    16819209    Defence
Chr09    7153856    7152515    LRR
Chr09    7230297    7229224    Defence    
Chr10    7737930    7736638    LRR-TM    
Chr10    7839150    7839821    LRR-TM        
Chr11    12440058    12440578    Defence    
Chr11    12716588    12716976    LRR    
Chr11    14534117    14533133    LRR-TM    
Chr11    17405732    17404812    LRR
Chr11    17545259    17544724    Defence

your help in this regard is highly appreciated

Thanks in advance

Ajay

[ajaybioinfo]

Dear Manual

Thanks for this prompt reply, this mean that i have to parse the data files contains 11 chromosome positioning R and DR genes using parse.assembly to generate the rice specific karyotype ?. and as a result it will generate all the necessary file to generate the corcos map is it so?

Kindly help me out as i am under high pressure of my supervisor to generate this map . hope you understand my situation

Thanks

[Sukhi Singh]

First of all, Manual is not a person. Secondly, you write the reply as a comment (under my answer). Then, read my answer again, you need to build the rice species karyotype (required by circos) using parse.assembly. Check the circos documentation page,

The first thing you will need is the karyotype for your genome. The karyotype defines the chromosomes (name, size, color) and, optionally, the cytogenetic band pattern. You can obtain karyotypes for many genomes from the cytoBandIdeo table (UCSC genome browser).

You can download the karyotype from the table browser or directly for human hg19 (Feb 2009) hg18 (Mar 2006), mouse mm9 (Jul 2009) mm8 (Mar 2006), rat rn4 (Nov 2004) rn3 (Jun 2003), or other species.

Circos is not a easy thing to get accustomed to quickly, clear this to your supervisor, that it will take time to produce what is essentially required.

[ajaybioinfo]

Dear Sukhdeep

Sorry for the error which i have made as i am very much new to bionformatics, and for late reply hope you understand :-)

actually i am facing problem in making the karyotype for the my data files. how can i use parse.aasebly cripts as my data contain the genes location on the respective chromosome like :- Chromosome Start End Function
Chr01 176373 175717 NBS-LRR
Chr01 228127 228513 Defence
Chr01 263322 262761 LRR-TM
Chr01 343320 345126 NBS-LRR
Chr01 476287 476105 LZ-NBS-LRR
................

second how can make karyotype for my genome is it possible that you can contact me at my mail id:- so that i can forward my data to you for creation of necessary file need to generate a circos map

i am very much thankful to you for your guidance and help

[ajaybioinfo]

Dear Sukhdeep

Please help me out i am waiting for your kind reply please suggest me the right way to make the said circos image

[Sukhi Singh]

I have updated my answer check

[Prakki Rama]

+1 "Manual is not a person".:-)

Answer 1

[PoGibas]

New R package is out - circlize; http://bioinformatics.oxfordjournals.org/content/30/19/2811.full.html?etoc

circlize implements circular visualization (http://circos.ca/) in R. Due to the natural born feature of R to draw statistical graphics, this package can provide a more general and flexible way to visualize huge information in a circular style

Nice visualizations + example code: http://jokergoo.github.io/circlize/

[Adam Cornwell]

This looks pretty good, now we have a we Circos-like options in R. Thanks!

[rob234king]

This looks great. You show examples of using it to generate images from matrix but if I wanted to show synergy using this example but with two genomes of 4 chromosomes, how would I generate the input? I'm not sure how the table creates the image and how to get to the table from the two genomes.

Answer 2

[Sukhi Singh]

From the manual, the data format is fine (looks like bed)

The line plot, scatter plot, heat map and histogram share the same data format.

#chr start end value [options]
...
hs3 196000000 197999999 71.0000
hs3 198000000 199999999 57.0000
hs4 0 1999999 28.0000
hs4 2000000 3999999 40.0000
hs4 4000000 5999999 59.0000
...

You just need another column pointing to the height of the histogram which is the value column. I have no idea what this file is, either convert it to a coverage/bedGraph file and you will have the 4th column or do the needful in collapsing/restructure the dataset to have this value.

For a genome wide plot, you would need to lay the karyotype of your species of interest and then overlay the histogram (the above modified file) on it. Check the example configuration. You will also be required to build the rice species karyotype as it doesn't exists by default. Check the script parse.assembly included in the Circos package for that.

Edit : How to make karyotype file for circos.

For making the karyotype file, use my script. My current working directory is circos/circos/data/karyotype within the circos installation where parse.karyotype script is present.

cat <(head -n1 chromosome.band.hg19.txt) <(curl ftp://hgdownload.cse.ucsc.edu/goldenPath/mm10/database/cytoBandIdeo.txt.gz| gunzip -c) > a && ./parse.karyotype a mm > karyotype.mouse.mm10.txt && rm a

You need to replace the mm10 ucsc link of cytoBand file with your organism of interest and you have the karyotype.

Cheers

[dimitrischat]

can anyone explain to me how do i get all the 4 columns (from bed files i dont have the 4th -value) ? for example from a bigwig file ?