TP53 META, A webtool to visualize effects of TP53 modulators and mutations on expression profiles with a focus on breast cancer

Compilation and comparison of multiple datasets that relate to a particular treatment can help us understand more about a scientific question of interest. Transcription factor protein 53 (TP53), also known as the “Guardian of the genome”, is one of the most mutated transcription factors in sporadic Breast Cancer. Even though there are multiple online databases that compile, annotate and list p53 targets, these gene lists can vary significantly and no web-tool is available to visualize them and perform a meta-analysis of their expressions. Herein, I have developed an R Shiny based tool called TP53 META that allows for comparison and visualization of up to four separate, two group RNA-seq datasets. TP53 META is able to normalize and perform differential expression analysis via limma or DESeq2 on different datasets, simultaneously. Moreover, it allows users to perform clustering, gene-set enrichment analyses, transcription factor-target enrichment, as well as visualization by using disease-gene and treatment-gene bipartite networks that can be generated for further understanding of individual or combination treatments. It is also possible to limit the use of TP53 META modules to a subset of the transcriptome, e.g., specific targets of p53 such as DREAM targets. TP53 META also calculates a meta-p-value using different available meta-analysis statistical methods and incorporates several public as well as in house datasets, most of which are of breast cancer cell lines, in which TP53 is modulated by siRNA treatments, inducers such as nutlin, doxorubicin, CHRNA5 depletion, or stable mutations in the heterozygous or homozygous background. Public datasets were acquired from GEO while for in house datasets, I have obtained the fastq files and converted to counts and integrated them on TP53 META web-tool and tested the following hypotheses to demonstrate the functionality of the web-tool: 1) Does CHRNA5 depletion, a known TP53 inducer, result in further downregulation of DREAM targets in MCF7 with the wildtype TP53 or heterozygous TP53 mutant MCF7 cells?; 2) Do TP53 inducers induce TP53 targets equally in wild type and mutant cancer cells? I also applied existing and novel methods to determine synergistic/antagonistic relationship of genes between the two treatments, e.g., doxorubicin alone or doxorubicin together with CHRNA5 depletion. Results showed that wildtype homozygous TP53 overexpression depletes DREAM targets more than heterozygous mutant MCF7 cells where a copy of mutant TP53 (R175H, R273H) exists. I have also found that CHRNA5 depletion in combination with doxorubicin can further regulate certain TP53 targets.