SEQUENCE ANALYSIS UNIX COMMANDS - Lecture 3

Getting genome sequence

1. Your hebrides account can't hold this much data, so only do this if you want the genome on your own computer.
2. The UCSC FTP site has the same hierarchy as UCSC Genome Bioinformatics downloads
3. FTP to a repository like UCSC
• ftp hgdownload.cse.ucsc.edu
• log in as 'anonymous' and give your email address when prompted
• cd goldenPath
• ls -F (and select a directory like hg17 [for the human May 2004 assembly])
• cd bigZips (for everything in a big file) or cd chromosomes (for separate files for each chromosome)
• get chromFa.zip (or whatever file you want)
• [continue for every file you want]
• quit
4. unzip chromFa.zip (or whatever file you want to unzip)

Mapping and extracting genomic sequence

1. First: format the genome using the faToNib command.
2. To BLAT search the genome with a multiple sequence fasta file,
• index the genome (get it into memory) with the gfServer command
• search the genome with the gfClient command.
3. Extract a region of a chromosome using the nibFrag command.
4. See Using BLAT on hebrides for all the details of these commands.
5. There is a command on hebrides called 'blat' but it only works to map one sequence to another but not to the entire genome.

Downloading the Ensembl annotation

1. FTP to Ensembl
• ftp ftp.ensembl.org
• log in as 'anonymous' and give your email address when prompted
• cd pub
• ls -F
• cd current_mart (or the directory you want)
• cd data/mysql (if you want EnsMart)
• cd ensembl_mart_28_1
• ls -F
• get ensembl_mart_28_1.sql.gz (or whatever version you want) to get a description of all the fields in all the files in this directory
• get hsapiens_ensemblgene_main.txt.table.gz (or whatever file you want)
• [continue for every file you want]
• quit
2. gunzip hsapiens_ensemblgene_main.txt.table.gz (or whatever .gz file you want to gunzip)
3. Since the file is tab-delimited, it can be parsed with Perl
4. The file can also be imported into a MySQL database
• using the CREATE TABLE command for the table(s) in the SQL file shown above, and
• and the LOAD DATA LOCAL INFILE command (see MySQL documentation for the details).

Querying the Ensembl database

1. Connect: mysql -u anonymous -h kaka.sanger.ac.uk (to get a MySQL prompt)
2. show databases;
3. use ensembl_mart_19_1; (or the database you want)
4. show tables;
5. describe hsapiens_ensemblgene_main; (or the table you want)
6. show tables;
7. enter a query:
• ex1: select * from hsapiens_ensemblgene_main limit 10;
• ex2: select * from hsapiens_ensemblgene_main where description like "%growth factor%";