BIOTECH-7005-BIOINF-3000

Assignment 6 - Transcriptome analysis [35 marks] Due Friday 01 November 2024

Due Friday 01 November

Your answers to all questions should be submitted to myUni as a .zip file containing four bash scripts, and a 5_final_assembly folder including files required in step 5, a screenshot file showing alternative splicing events in Step6, and a text file inlcuding your answers to theoretical questions. [1 mark]

The .zip filename must start with your student number and your bash script must be able to run without errors. Meaningful comments are strongly advised [1 mark]

For all scripts, please use the directory ~/Assignment6 as the parent directory for all downloads and analysis. You will be expected to hard code this into your scripts. Use an organised folder structure to store files generated in this assignment. [1 mark]

Assignment6/
├── data
├── DB
├── results
│   ├── 1_QC
│   ├── 2_clean_data
│   ├── 3_denovo_assembly
│   ├── 4_genome_guided_assembly
│   └── 5_final_assembly
└── scripts

Practical questions [28 marks in total]

Step 1. Write the first script to:

• Download the genomic sequence (i.e. fasta file) and annotation (i.e. gff file) of the model plant Arabidopsis thaliana to the subdirectory DB from the Ensembl ftp directory as you did in assignment 4 [1 mark]
• Download the BUSCO “brassicales” lineage dataset using this link. Store the lineage dataset in the DB subdirectory and decompress and untar it [1 mark]
• Build the genome index for GMAP [2 marks]
• Build the genome index for STAR [2 marks]

Step 2. Write the second script to copy the sequencing data to your data directory and carry out QC. The data are in ~/data/Transcriptomics_data/Assignment/. [1 mark] Include the following steps:

• Get QC report for raw reads [1 mark]
• Trim adaptor and low-quality sequences using cutadapt [1 mark]
• Get QC report for clean reads [1 mark]

Step 3. Write the third script to:

• Do de novo transcriptome assembly for clean reads from chromosome 4 using Trinity [3 marks]
• Get the genome coordinates file (GFF3) of de novo assembled transcripts using GMAP [1 mark]
• Assess the de novo assembly quality using BUSCO [2 marks]

Step 4. Write the fourth script to:

• Do genome guided transcriptome assembly for clean reads from chromosome 4 with STAR and StringTie [3 marks]
• Get a fasta file of genome guided assembled transcripts using gffread [1 mark]
• Assess the genome guided assembly quality using BUSCO [2 marks]

Step 5. Organise the final results:

• Copy QC reports for raw reads (two html files) and clean reads (two html files) into 5_final_assembly [1 mark]
• Copy de novo assembly results (including one fasta file and one GFF3 file) into 5_final_assembly [1 mark]
• Copy genome guided assembly results (including one GTF file and one fasta file) into 5_final_assembly [1 mark]

Step 6. Identify one assembled transcript including alternative splicing events with read evidence (Take a screenshot from IGV and include the screenshot file in you assignment submission) [3 marks].

Theoretical questions [4 marks in total]

1. Can we get all genes assembled from one transcriptome sequenced from a sample collected from one particular tissue at one particular developmental stage? [1 mark]
2. Give the reason(s) supporting your answer in quesion 1 [3 marks]

Resources (also in assignment 4)

• The genome (fasta) for the model plant Arabidopsis thaliana can be found here: http://ftp.ebi.ac.uk/ensemblgenomes/pub/release-51/plants/fasta/arabidopsis_thaliana/dna/ and the annotation (gff3) files can be found here: http://ftp.ebi.ac.uk/ensemblgenomes/pub/release-51/plants/gff3/arabidopsis_thaliana/. There are multiple files in these directories, so take care to choose the full genome sequence and full genome gff files.
• The genome file should be “Arabidopsis_thaliana.TAIR10.dna.toplevel.fa.gz” and not contain an “rm” or “sm” in the filename. These refer to the genome being “repeat-masked” i.e. all the repetitive elements have been ignored, or “soft-masked” were the repeats and low-complexity regions are in lower-case letters (acgt not ACGT)