Chapter 9 Gene Prediction Methods

Objectives

  1. Annotate assembled genomes
  2. Interpret and summarize annotation output files

After assembling the genome, we need to identify which regions are genes. This is called gene annotation or gene prediction, and uses a combination of ab-initio approaches with homology-based searches to predict and annotate the genes.

We will use prokka, a state-of-the-art program for prokariotic genome annotation. This program uses both approaches to predict genes.

9.1 Installing prokka

#*# Installing miniconda

Before we can start installing prokka, we need to create a computational environment for it to be able to run correctly. Miniconda provides that environment by creating a python instance that groups all the requirements for these programs to run.

bash
/Smaug_SSD/bin/Miniconda3-latest-Linux-x86_64.sh

Say yes to all the prompts, and let miniconda install.

Then, close your window and reconnect to Smaug.

9.1.1 Installing the prokka environment and program

After logging back in, we need to install Prokka using miniconda as following:

bash
conda init bash

log out and log in again, and then:

bash
conda create -n prokka_env -c conda-forge -c bioconda prokka
conda activate prokka_env

We can finally run prokka. Test it by running

prokka

A long prompt shoud appear.

9.2 Annotating the genomes using prokka

Usually, these computationally complex programs are very long and cumbersome to run, but the develkopers of prokka have made it super easy to run.

Within your sample folder, run the following code:

prokka --dbdir /home/jtabima/micromamba/envs/prokka_env/db --genus Synechoccocus --strain the_name_of_your_strain --outdir your_sample_name_annotation --prefix your_sample_name  your_assembly.fasta

And let it run!

9.2.1 Summarizing the results from prokka

After prokka is done, the outputs should be in your your_sample_name_annotation folder.

  1. Using the head command, answer the following question:
  1. Check the .fna, .fsa, .ffn, .faa, .gff, .tbl, .tsv, and .gbk files and explain what are their contents below:
  • .fna file:
  • .fsa file:
  • .ffn file:
  • .faa file:
  • .gff file:
  • .tbl file:
  • .tsv file:
  • .gbk file:

Hint: Check the Prokka manual

  1. Using your knowledge on the command line, can you count the number of predicted proteins, and all the prediction transcripts (CDS, rRNA, tRNA, tmRNA, misc_RNA)

  2. Open the .txt file. Answer the following questions:

  • How many CDS are found? Do they match your counts from the previous step?
  • Define rRNA, tRNA and tmRNA (cite your references please) and present the counts of each RNA type.
  1. Using the head command, describe the columns of your .gff file.
  1. Use the following code to create a summary table of your functional annotations:
cut -f 7 your_sample_name.tsv | sort | uniq -c | sort -nr > gene_counts.txt

Open the gene_counts.txt file and answer the following questions:

  1. What is the most abundant gene prediction?

  2. What is the most abundant gene prediction that is not a hypothetical/predicted protein?

  3. What are the most common five gene annotations in your genome?

  1. Finally, select three random hypothetical proteins from your faa file. Run them in InterPro Scan (https://www.ebi.ac.uk/interpro/search/sequence/).

Answer the following:

  1. Did any of these proteins have an annotation? Add it below.

  2. Add the path of your faa file here: