This is a test version of Biostars. For the public version, visit https://www.biostars.org.
Where do I start with Circos plot?

Hi, I've been looking for helps with Circos plot and I couldn't find the very very basic helps.

So, I have about 20 short genome sequences (~30K) that I want to show their similarity and simple genome information. I found out that circos plot is the perfect visualization of doing so and I searched tutorials and manuals.

However, I realized that all tutorials and manuals assume that I already have a table that is ready to be manipulated with. But the thing is.. I don't have a table. I don't know what parameters I need to put into the table to show their comparison. I have no idea how to generate comparative data to construct a table for Circos plot.

It is like.. everyone is so kind enough to teach me how to run; from how to put on a pair of shoes, tie up the shoe laces, how to take a first step, from pinky to large toe, and how to run fast. But here I am, wondering.. so where do I get a pair of shocks and shoes?

I know that I look dumb.. but I seriously need some help here.... please save me

Thank you in advance..!!

circos genome comparison
• 14,561 views • link
updated by Andrewoods Andrewoods • written by krmoon518 • krmoon518

I tried drawing synteny maps with "EasyFig" and genbank files before and that's pretty much all that I know about visualization of comparing two for more genomes. Hope that explained my state of ignorance....

• 0 views • link
written by krmoon518 • krmoon518

Be careful with circos plots. They are unreadable most of the times. People have strong opinion about them.

• 0 views • link
written by poisonAlien poisonAlien

Thank you for the advice. I will keep it on my mind!

• 0 views • link
written by krmoon518 • krmoon518

7 answers

I have used this tutorial in past to get started with Circos: http://pbgworks.org/sites/pbgworks.org/files/Introduction%20to%20Circos.pdf

• 0 views • link
written by Sej Modha Sej Modha

Thank you for the tutorial! The tutorial you showed me tells me to type in.. <plot> type=scatter file=data/tmlong/pc1/pmpc1.txt r1=0.71r

and.. this was my problem. I do not have the pmpc1.txt (or anything that is equivalent of). My question started from here. Every tutorials assume that I already have a table made to be inputed. However, as a beginner, I did not have any table or information to put in. I didn't know if I was to add e-values or percent similarities to the table, and didn't know in what order. Please don't get me wrong, your comment was very nice but I'm simply not ready for that yet.. Thanks anyway!

• 0 views • link
written by krmoon518 • krmoon518

I believe that page 35-36 provides more info about the pmpc1.txt file you're looking for.

• 0 views • link
written by Sej Modha Sej Modha

This question has been answered previously. Please use the search tool next time. Here is one of the posts to answer your question: C: how to make a OmicCircos config files

• 0 views • link
written by theobroma22 theobroma22

Thank you for the link. I just followed the link and read it and I kind of see that the table could be composed of the name/genome start/genome end/and so on. So.. after that, to show the relationship between the genomes that I want to display, (draw ribbons as many of example pictures depicted), what kind of "comparative" values would I need to add?

• 0 views • link
written by krmoon518 • krmoon518

This part is contained in the link denoted code and objects like link.i, link.j, link.df, link.c, etc. For genomes, drawing ribbons is typically from one gene start position on Chromosome A to another gene start position on Chromosome Z, but you can also use the average start position if that's what you choose to use for your input.

• 0 views • link
written by theobroma22 theobroma22

Do people normally type those gene positions into the table? like.. one-by-one? (I am so sorry to bug you with such toddler questions)

• 0 views • link
written by krmoon518 • krmoon518

It can be done by hand or programmatically. You may need to copy and paste from a file or files to make the link file or you can make this programmatically. So, there is more than one way to do this.

• 0 views • link
written by theobroma22 theobroma22

Try using Sibelia. It's a synteny calculation tool, but it uses Circos to generate its figures. You can use their config files to get started, if you don't actually want to use Sibelia itself (though it sounds like it might work nicely for what you're doing).

• 0 views • link
written by Joe Joe

Thank you so much!!! although I am still not clear how the table got constructed, I got the picture that I wished to have! Now I will try to understand the steps and modify the pictures! Thank you so much!

• 0 views • link
written by krmoon518 • krmoon518

Sibelia does calculate the synteny for me and allows me to construct circos plot. However, I realized that no more than one ribbon is drawn from one gene to the other. In my data, I found 20 genomes that are syntenic to a genome A (all 23 were searched based on genome A) So, I predicted that multiple ribbons would be drawn out to each of the 20 genomes but that did not happen. Instead, only the most solid ribbons were drawn, leaving my genome A to be poor of ribbons. I apologize for too much unnecessary comments. I'm in the middle of searching ways to solve the problem and I wondered if .. you would know.. Thanks again..!

• 0 views • link
written by krmoon518 • krmoon518

Have you experimented with the fine/loose parameter and the synteny block size?

• 0 views • link
written by Joe Joe

sorry for the late reply. Yes I just did, and loose parameter certainly improved my data. Thank you!

• 0 views • link
written by krmoon518 • krmoon518

Great, if one of the answers here has resolved your question, don't forget to accept one of them

• 0 views • link
written by Joe Joe

Also for genome comparisons BRIG is a fairly common choice.

https://bmcgenomics.biomedcentral.com/articles/10.1186/1471-2164-12-402

• 0 views • link
written by Joe Joe

Hello there.

If you are not used to R or programming, you can begin with Excel to get such a table.

First, what you need is to have a 'starting table' where you can have two or three columns. For a group of genomes (I suppose you want to show what genes are shared among them and which ones are not), you can have a first column where you can add the name of the genome. On the second one you can have the name of the specific genes. And if you want a third one (I am not in genomes analyses at all), maybe you can have the number of copies of these genes in each genome. You may have something similar to this:

Genome    Gene    No. copies
Genome1   gene1    2
Genome1   gene2    1
Genome2   gene1    3
Genome2   gene3    5
...

Then, when you have this table you can create a 'Dynamic Table' with these data to get a table where you can see the shared and unshared genes between the genomes.

From this point onwards, creating the CIRCOS diagram shoudn't be so hard. Take into account that you can run CIRCOS on its website or on your computer. If you download the executable file you can change a lot of parameters just by following the files that come with it. Just read them carefully and follow each instruction on them and you will get gorgeous graphs.

Mines:

https://www.researchgate.net/publication/316189005_Microbial_succession_dynamics_along_glacier_forefield_chronosequences_in_Tierra_del_Fuego_Chile

• 0 views • link
written by mafernandez • mafernandez

Here is tutotrial, but for the last steps, I can't find the perl scripts. Does anyone know where there are?

• 0 views • link
written by Ric Ric

Maybe you want to try shinyCircos. It's an interactive APP. https://github.com/venyao/shinyCircos http://shinycircos.ncpgr.cn/

• 0 views • link
written by Andrewoods Andrewoods

Log in to answer this question.

More posts like this