02-BLAST+ at the terminal

Learning Outcomes

• Use of BLAST+ at the terminal
• Getting help at the command line
• Building custom local BLAST+ databases
• Understanding the BLAST+ command-line
• Getting output in multiple formats
• Interpreting BLAST+ output

Table of Contents

1. Introduction
2. Using BLAST+ in the terminal
   1. Build a BLAST+ database
   2. Exercise 01
   3. Construct a BLASTN query
   4. Exercise 02

Introduction

The BLAST/BLAST+ package can be installed on your own machine (desktop or laptop) or on a shared server. This gives you full control over how to use the program, and allows you to build custom databases (useful for proprietary information). However, you are limited to the computing power you have available. Happily, BLAST doesn't require excessive amounts of computing resources and for many tasks a desktop or laptop machine is sufficient.

Resources

• ncbi-blast+ download
• Original publication: Altschul et al. (1990)80360-2)
• Gapped BLAST publication: Altschul et al. (1997)

Using BLAST+ in the terminal

• If necessary, open a terminal window in the virtual machine (VM), or a new tab in your terminal window.

• Change directory to the ~/Teaching-IBioIC-Intro-to-Bioinformatics/02-sequence_databases lesson directory:

cd ~/Teaching-IBioIC-Intro-to-Bioinformatics/02-sequence-databases
ls

• Establish that BLASTN works by issuing a command to get the short help message:

blastn -h

Build a BLAST+ database

The program that builds BLAST+ sequence databases is makeblastdb. You can get basic help on the command by issuing:

makeblastdb -h

To build a `BLAST` database we need to provide the following information:

1. A file containing the sequences that will be in the database
2. What kind of sequence (nucleotide or protein) data the file contains
3. A name for the database (optional)
4. A path to write the database files to (optional)

• Create a new BLAST database with the following command:

makeblastdb -in data/kitasatospora/GCA_001905465.1_ASM190546v1_cds_from_genomic.fna \
            -dbtype nucl \
            -title kitasatospora_cds \
            -out data/kitasatospora/kitasatospora_cds

This will return some information to the terminal, and create the database.

This creates three files, which together comprise a new BLAST nucleotide database against which you can make queries.

Exercise 01: Get BLAST help at the Terminal

Use the following commands to get the long-format help messages for BLASTN and BLASTX:

• `blastn -help`
• `blastx -help`

Pay particular attention to the options for output `-outfmt` and `-out`, and the options that control the general search parameters.

Construct a BLASTN query

After looking at the help information in the exercise above, you will have seen that there are several relevant input options:

• -query: path to the query sequence(s)
• -db: path to the BLAST database
• -outfmt: the output format you want BLAST to produce
• -out: path to the output file you want BLAST to write

Building a `BLAST` query at the command-line/terminal is a matter of using the appropriate program (here, `blastn`) and passing it the input options you need to use.

In this case:

• your query sequence is data/kitasatospora/lantibiotic.fasta
• the database you're searching against is the one you created above: data/kitasatospora/kitasatospora_cds

and you'll generate output in two formats (the same ones that were produced from the NCBI website search). You will need to construct two commands, each with the same query and database, but different output format values, and output filenames:

• no format specified, filename: output/kitasatospora/terminal_blastn_query_01.txt
• format: 6 (tabular), filename: output/kitasatospora/terminal_blastn_query_01.tab

• Run the first command at the terminal:

blastn -query data/kitasatospora/lantibiotic.fasta \
       -db data/kitasatospora/kitasatospora_cds \
       -out output/kitasatospora/terminal_blastn_query_01.txt

The command will run without producing any output on the screen, but you can see the first few lines of the output by issuing:

head -n 40 output/kitasatospora/terminal_blastn_query_01.txt

• Run the second command, now specifying a different (tabular) output format:

blastn -query data/kitasatospora/lantibiotic.fasta \
       -db data/kitasatospora/kitasatospora_cds \
       -outfmt 6 \
       -out output/kitasatospora/terminal_blastn_query_01.tab

You can inspect the contents of this file by issuing the command:

less output/kitasatospora/terminal_blastn_query_01.tab

QUESTIONS

1. How many hits were found
2. How large was the database?
3. How does the tabular output compare to the plain text output?

Exercise 02: Using BLAST at the Terminal

Using BLAST in the terminal:

• Conduct a `BLASTX` query with `data/kitastaospora/lantibiotic.fasta` against the `data/kitasatospora/kitasatospora_proteins.faa` database, writing results in `Text` and `Tabular` format to:
  • `output/kitasatospora/terminal_blastx_query_02.txt`
  • `output/kitasatospora/terminal_blastx_query_02.tab`

QUESTIONS

1. How many hits do you find?
2. What is the "best hit" to the query? Why do you think it is the "best hit" (what in the results tells you this?)
3. At what point do you think the matches start to become less reliable? Why do you think this? (*HINT:* inspect the alignments)