Shulou(Shulou.com)06/02 Report--

This article introduces how to use Python to draw Manhattan map and QQ map in GWAS analysis, the content is very detailed, interested friends can refer to, hope to be helpful to you.

Manhattan map and QQ map are the two most common graphics in genome-wide association (GWAS) analysis, which are basically standard in GWAS and can be seen in almost every GWAS article. I also explained their role and message in my last blog post about GWAS, and here we only focus on how to present it effectively with Python and geneview.

First of all, prepare some data as examples.

The data I use here is the GOYA, a Danish GWAS study on obesity among young people in 2011, and the data are also obtained from their published results, with a total of 5373 samples, including 2633 overweight individuals (case) and 2740 normal individuals (control), which is not bad in terms of sample size. In order to make it easy to use, I processed it, including the generation of PED and MAP files to GEN files, and repeated the correlation analysis of case-control to calculate the significant degree of obesity related to each SNP locus studied on the chip (that is, p-value). Finally, the resulting data was extracted into a data set, which can be downloaded here (15.6Mb CSV format).

[note] although the above mentions some domain terms and related file formats, please don't bother if you don't understand, because the subsequent processing is based on this final data set.

Next, you need to add the geneview package to your Python. There are many different ways, but it is recommended to use pip directly. The following is a more stable release version, which can be entered directly at the terminal command line (Linux or Mac):

Pip install geneview

Alternatively, you can install the version under development directly from github:

Pip install git+git://github.com/ShujiaHuang/geneview.git#egg=geneview

The third way is to download the source code directly and then compile it by yourself. Although this is not recommended (because there are dependent packages that must be downloaded and installed on your own, the process will be troublesome and inefficient), it can only be done for some clusters that cannot connect to the external network. All three methods are feasible.

Manhattan chart

Download the sample data:

Wget https://raw.githubusercontent.com/ShujiaHuang/geneview-data/master/GOYA.csv

First take a brief look at the format of the data:

ChrID,rsID,position,pvalue

1,rs3094315,742429,0.144586

1,rs3115860,743268,0.230022

1,rs12562034,758311,0.644366

1,rs12124819,766409,0.146269

1,rs4475691,836671,0.458197

1,rs28705211,890368,0.362731

1,rs13303118,908247,0.22912

1,rs9777703,918699,0.37948

1,rs3121567,933331,0.440824

There are 4 columns (separated by commas), which are: [1] chromosome number, [2] SNP rs number, [3] locus position on chromosome, [4] significant difference (pvalue). In this Manhattan diagram, we only need to use columns 1, 3, and 4, while the QQ diagram only needs column 4-- pvalue.

Let's start by drawing a map of Manhattan. Let's first read the data we need into a list, and we can do this:

Import csv

Data = []

With open ("GOYA.csv") as f:

F_csv = csv.reader (f)

Headers = next (f_csv)

Data = [[row [0], int (row [2]), float (row [3])] for row in f_csv]

Now the data in GOYA.csv is stored in the data list, because Python reads the data in the file, it is stored in the string type, so it is necessary to do some type conversion for the data in the third and fourth columns.

Next, call the Manhattan chart function in geneview.

Import matplotlib.pyplot as plt

From geneview.gwas import manhattanplot

Ax = manhattanplot (data, xlabel= "Chromosome", ylabel= "- Log10 (P-value)") # this is the function of Manhattan plot

Plt.show ()

It only takes such a sentence of code to create a beautiful Manhattan map. It is necessary to point out again that geneview is developed on the basis of matplotlib, and the created graphic objects actually still belong to many custom graphic styles within matplotlib,geneview. At the same time, it encapsulates a large number of chart types that only belong to genomic data, but the output format and interface display of the graph are still the same as matplotlib. So here we use matplotlib.pyplot 's show () function (in the example above: plt.show ()) to show the Manhattan diagram we have drawn. If you want to save the drawing, simply execute `plt.savefig ("man.png") `, which will generate a png map of Manhattan called "man.png" in this directory. If you want to save it in pdf format, simply change the suffix of the file name you want to save to ".pdf" ("man.pdf"). The following formats are supported: emf, eps, pdf, png, jpg, ps, raw, rgba, svg, svgz, etc. As for the latest ones, please refer to the matplotlib documentation.

In addition, each drawing function in geneview is flexible enough, and we can make some adjustments according to our own needs, such as:

Xtick = ['1ZHANGZHANJI' 2ZHANGZHANGZHANGZHANG: 5ZZHANGZHANG: 5ZHANZHANG: 6ZHANZHANG: 7ZHANZHANG: 8THANG: 9THERONING 10: 11: 12: 12: 13: 14: 16: 18, 22: 22]

Manhattanplot (data

Xlabel= "Chromosome", # set the x-axis name

Ylabel= "- Log10 (P-value)", # sets the y-axis name

Xtick_label_set = set (xtick), # defines the scale display on the horizontal axis

# set the size of the scatter points in the figure

Alpha=0.5, # adjust scatter transparency

Color= "# f28b1e recording 9a0dearecoverye0dccrecovery63b8ff", # setting a new color combination

)

Realize the new color combination, limit the scale display on the x-axis and adjust the size of the scatter. You can even change scatter points to lines:

Manhattanplot (data

Xlabel= "Chromosome", # set the x-axis name

Ylabel= "- Log10 (P-value)", # sets the y-axis name

Xtick_label_set = set (xtick), # defines the scale display on the horizontal axis

Alpha=0.5, # adjust scatter transparency

Color= "# f28b1e recording 9a0dearecoverye0dccrecovery63b8ff", # setting a new color combination

Kind= "line"

)

For other adjustments, please see the relevant instructions in the geneview documentation.

QmurQ diagram

The qq diagram only needs to use the pvalue column in the example above:

Import csv

Import matplotlib.pyplot as plt

From geneview.gwas import qqplot

Pvalue= []

With open ("GOYA.csv") as f:

F_csv = csv.reader (f)

Headers = next (f_csv)

Pvalue = [float (row [3]) for row in f_csv]

Ax = qqplot (pvalue, color= "# 00bb33", xlabel= "Expected p-value (- log10)", ylabel= "Observed p-value (- log10)") # Qmuri Q chart

Plt.show ()

Similarly, you can also adjust the map according to your own needs.

The above is how to use Python to make Manhattan and QQ diagrams. The integration function of geneview simplifies such a process.

In addition, if you have also read the GOYA study of the Danes, you will find that the two pictures above are basically the same as those in the article. Of course, I have done some data cleaning by myself, and the results will still be a little different. Although it is still a bit early to draw conclusions, on the whole, I should also be able to repeat the results smoothly through this data set.

Finally, attach the code to draw Manhattan and QQ diagrams using geneview:

(1) Manhattan chart:

(2) QQ diagram:

On how to use Python to draw Manhattan and QQ maps in GWAS analysis is shared here, I hope the above content can be of some help to you, can learn more knowledge. If you think the article is good, you can share it for more people to see.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.