USING R PROGRAMMING Choose your own adventure RNAseq analysis For this homework, the goal is to find an RNAseq counts table and sample metadata. Develop a plan to test a hypothesis through differenti

USING R PROGRAMMINGChoose your own adventure RNAseq analysisFor this homework, the goal is to find an RNAseq counts table and sample metadata.  Develop a plan to test a hypothesis through differential expression to compare 2 groups.  Perform the steps of a differential expression and functional enrichment analysis. Present your results.Step 1:  Find an RNAseq dataset you can use.  You want an RNAseq dataset with counts.  Describe the dataset.  Where did it come from? What is the disease?  Can you comment on what steps were used to analyze the data to this point? Summarize some relavent metadata (metadata = clinical phenotypes)Step 2: Select your phenotype to compare—split the data into 2 groups.  Do you want to compare tumor vs normal in the case of cancer?  Old patients vs young patients? How would you split a continuous variable in to 2 groups? Maybe you could compare something to do with severity like Higher Grade vs Lower grade?  Maybe combine grades I and II to low and grades III and IV to high?Step 3: What do you think will happen?  Any prior knowledge you can find to predict a biological difference?  Like “I expect KRAS to be differentially expressed in a comparison of Lung cancer tumor compared to adjacent normal” or “I expect genes relating to immune response to be different comparing patients with active flu infection compared to healthy patients.” Define your limits for significance: “For a gene to be significant, I require FDR<0.05” or “For a gene to be significant, I require FDR<0.1 and abs(log2FC)>1”Step 4: What is your sample size by phenotype?  Do you expect your phenotype to have a large effect on gene expression?  Plot the samples you are comparing on a PCA plot colored by your phenotype and make some inferences.Step 5: Do differential expression analysis.  Use deseq2 for the easiest approach.  For a challenge, try edgeR or limma (look for tutorials).  Describe how the counts are normalized and what the differential expression analysis is doing. Step 6: Generate DE summary plots of volcano and heatmap of significant genes.  How well did your analysis meet your significance limits? Modify those limits if you need to in order to make a heatmap. For example if no genes met your FDR<0.05 threshold, plot the top 50 genes with the smallest p value.  If this happens, why do you think nothing was significantly different?  Is your chosen set of differentially expressed genes a good panel for differentiating your groups?Step 7: Taking the genes that meet your significance limits (or lowered limits to use at least 200 genes) do functional enrichment analysis and describe the results.  If you find a significant functional enrichment, display a plot to summarize the results.  Do these results make sense? If you did not find a significant enrichment, why not?Put your code and comments and plots in an Rmd and knit the report into an html.  Upload the Rmd and the html.  If this is a challenge, put your comments and plots in a word doc and upload R code and word doc summary.