In my last post, I illustrated how the Kaplan-Meier estimator can be used to estimate the survival curve of mRNA half-lives. In this post I will expand on that analysis and show how to compare two mRNA half-life Kaplan-Meier curves, each corresponding to a measured gene outcome, to see if mRNA half-life differs between outcomes. (Sorry, I can’t reveal what the “outcomes” are!).

Example source data is shown in the following table. The “half_life” variable indicates the measured half-life of each gene’s mRNA [1], while “event_observed” indicates whether the event (half-life reached) was observed during the measurement experiment. We can see that one of the cases in the example source data below is right censored, i.e., the half-life of the mRNA was not reached by the time the experiment ended. The “group” variable indicates which outcome each gene’s mRNA corresponds to:

Plotting the Kaplan-Meier estimate curves for each outcome yields:

We see that the median half-lives do indeed look different on the graph, and in R’s estimate of the median values:

But is the difference statistically significant? We can answer that question with the Mantel-Cox log rank test, shown on the last line of the Cox proportional hazards regression screenshot below:

Here the p-value is 1.308e-09, so we reject the null that the curves are equal. We therefore conclude that mRNAs associated with outcome B generally have a longer half-life.

# References

1. Lioudmila V. Sharova, Alexei A. Sharov, Timur Nedorezov, Yulan Piao, Nabeebi Shaik, and Minoru S.H. Ko. **Database for mRNA Half-Life of 19 977 Genes Obtained by DNA Microarray Analysis of Pluripotent and Differentiating Mouse Embryonic Stem Cells.** DNA Res (2009) 16(1): 45-58 first published online November 11, 2008 doi:10.1093/dnares/dsn030