OncoLnc: A straightforward Tool

OncoLnc Web Tool: Interactively exploring survival correlations, and for downloading clinical data coupled to expression data for mRNAs, miRNAs, or lncRNAs from TCGA.

In Brief

• Offical website and Publication Link.

• Data Wrangling and Re-analysis with R Package.

• Summary

OncoLnc Usage

• Enter Interested Gene, DONSON, and click on Submit.

• Choose Interested cancer, KIRC, and click on Yes Please! of the cancer. Herein, DONSON gene ranked in #1.

• Setup value for Lower Percentile and Upper Percentile with the purpose of dividing the patients without overlapping slices.

• Click on Submit.

• Click on Click Here, to get the excel file of this data.

Re-analysis with R

Reading Data and Check Basic Information

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14

a=read.table('KIRC_DONSON.csv', header = T, sep = ',', fill = T)

colnames(a)
output
[1] "Patient"    "Days"       "Status"     "Expression" "Group"  

table(a$Status)
output
Alive  Dead
  151   109

fivenum(a$Expression)
output
[1]  15.780  81.605 155.420 266.310 905.630

Load Packages/Library and Make Basic Plot

1
2
3
4
5
6
7
8

library(ggstatsplot)
library(ggplot2)
library(survival)
#install.packages('survminer')
library(survminer)

# 'Dead':1; 'Alive':0
a$Status=ifelse(a$Status=='Dead',1,0)

Expression vs Group

1
2

ggbetweenstats(data=a, x=Group, y=Expression)
ggsave('ExpressionGroup.png')

Basic Survival Curve

1
2
3
4
5
6

fit = survfit(Surv(Days,Status)~Group,data = a)
fit
summary(fit)

ggsurvplot(fit,conf.int = F, pval = TRUE)
ggsave('survival_R.png')

Survival Curve With More Information In Figure

1
2
3
4
5
6

#  Better Display
ggsurvplot(fit,palette = c("#E7B800","#2E9FDF"),risk.table = TRUE,pval = TRUE,
conf.int = TRUE,xlab="Time in months",
ggtheme = theme_light(),
ncensor.plot=TRUE)
ggsave('survival_R1.png')

Herein, R.script is reused to make survival curves.

Summary

Overall

• The code was run with Python 2.7.5, NumPy 1.7.1, and rpy2 2.5.6.

• It can require upwards of 6GB of RAM.

• OncoLnc runs on Django 1.8.2, Python 2.7, matplotlib 1.2.1, NumPy 1.7.1, rpy2 2.5.6, uses the SQLite3 database engine, and utilizes Bootstrap CSS and JavaScript, and Font Awesome icons.

Reproduce information

• Put the expression files to the correct locations.

• Run the desired Cox regressions.

• Go to the cancer of interest and run the cox_regression.py file from the command line.