Your task will be to analyse the data in a three$-$step pipeline and then focus on the interpretation

of results in step $4$ :

Step $1$: For each variety $(x$ and $Y),$ combine the gene expression data and process into a

single file $(12\%).$

Step $2$: Perform initial processing to get the gene expression data into a usable format $(12\%).$

Step $3$: Manipulate and visualise the data to address the aims of the study $(24\%).$

Step $4$: Complete an MCQ based on your results and their biological interpretation $(12\%).$

Input Files

You will see that the Unit$8/$Matrix$/$ directory contains the following files and folders you will need

for this assessment: The gene expression data is currently split up into several files within various subdirectories of the $/$rds$/$homes$/$u$/$username$/$module$-1-$intro$-$to$-$biology$-$and$-$programming$/$Unit$8/$Matrix directory. Your task will be to analyse the data in a three$-$step pipeline and then focus on the interpretation

of results in step $4$ :

Step $1$: For each variety $(x$ and $Y),$ combine the gene expression data and process into a

single file $(12\%).$

Step $2$: Perform initial processing to get the gene expression data into a usable format $(12\%).$

Step $3$: Manipulate and visualise the data to address the aims of the study $(24\%).$

Step $4$: Complete an MCQ based on your results and their biological interpretation

Write a Bash script that can be run within the Matrix$/$directory to process the gene expression matrices to produce the following two output files:

A single csv file containing the gene expression matrix for variety X$,$ including the header line. You should name this file GEM$\_$X$.$csv $($GEM stands for Gene Expression Matrix$).$

A single csv file containing the gene expression matrix for variety Y$,$ including the header line. You should name this file GEM$\_$Y$.$csv

In the file headers, each sample name for the columns of the matrix currently has the following format:

Variety, followed by condition code $($C$,1,2$ or $3)$ and biological replicate $($a$,$b or $$c$).$ C is the control, and $1-3$ are the stress conditions. For example,

denotes Variety X$,$ Treatment C and replicate $$a$.$

Your gene expression matrix files should:

Contain only unique genes located on the $12$ chromosomes, with genes sorted by chromosome.

Change the column labels in the header to a more convenient format, replacing $$a with Rep$\mathrm{.}1,-$b with Rep$\mathrm{.}2$ and $$c with Rep$\mathrm{.}3.$ In the above example,

becomes

$.$

Your submitted Linux script should be called Linux$\_$your initials.sh $($e$.$g$.$ Linux$\_$LC$.$sh$)$

Now you have collected all the data into a single gene expression matrix for each Variety, you should use Python to select the gene expression data from two conditions only and change the format into something that will be easier to analyse in R$.$

Create a Jupyter Notebook or a plain Python script which takes the following inputs:

The two gene expression matrices you created in Step $1$ for varieties X and Y$.$

The two files containing information on the differentially expressed genes.

And gives the following outputs:

Two gene expression matrices $($one for each Variety$),$ each with a header, and containing gene expression from two conditions only: the control condition $($in columns $2-4$; replicates $1-3)$ and stress treatment condition $1($columns $5-7$; replicates $1-3).$ You should name your files all$\_$VarX$\_$TwoTimePoints.csv and all$\_$VarY$\_$TwoTimePoints.csv$.$

Two files, each with a header, containing information on the differentially expressed genes for each variety, and containing a subset of columns in this order: gene$\_$name, log$2$FoldChange, padj, Athaliana$\_$geneID and Gene$\_$Function, followed by an additional six columns for the expression of the three replicates in each of the two conditions in the order given above. You should name your files Leaf$\_$DEGs$\_$VarX.csv and Leaf$\_$DEGs$\_$VarY.csv$.$

The meaning of the selected columns relating to differential gene expression are given below:

log$2$FoldChange: The logarithm to the base $2$ of the expression ratio between the treatment condition and the control condition. This is a convenient way to quantify the change in expression between two conditions because a log$2$FC $=1$ means the expression has doubled in the treatment condition whereas log$2$FC $=-1$ would mean the expression has halved.

padj: An adjusted p$-$value. You will learn more about the p$-$value in a future module. For now, this probability value gives a measure of how likely it is that the gene expression change is "real" instead of being seen by chance. The lower the p$-$value, the more "significant" we deem the change in expression to be$.$

Athaliana$\_$geneID: The corresponding gene in the model plant Arabidopsis thaliana. This can be very useful information for downstream functional analysis of genes when working with crops like potatoes.

Gene$\_$Function: Functional information available for the gene $($what does it do$?).$