Merge branch 'master' of github.com:swcarpentry/2012-11-scripps

SSH/blast.rst

@@ -0,0 +1,115 @@
+Installing and running standalone BLAST 
+
+BLAST - Basic Local Alignment Search Tool
+
+
+****  Download the BLAST program ****
+Use an FTP client or Firefox and go to 
+ftp://ftp.ncbi.nih.gov/blast/executables/
+
+Click on 'LATEST'
+
+For Macs you can use the .dmg file
+Scroll to the bottom of the page and click on or transfer the *.dmg file
+Right now that file is
+ncbi-blast-2.2.25+.dmg
+
+Go find that file and double click it
+A folder will open with the ncbi-blast-2.2.25+.pkg file in it
+  double click on that file and click through the steps to install it
+
+It's installed!
+
+Modify the .ncbirc file that was just created in your directory to have it point to the place where you'll be creating your databases
+
+Open a terminal and at the command line type
+
+pico .ncbirc
+
+Make the changes
+e.g.
+
+[BLAST]
+BLASTDB=/Applications/ncbi-blast-2.2.24+/db
+
+
+To save and exit type
+Ctrl-X
+when it asks you if you want to save, type Y and then Enter when it asks if you want to save to .ncbirc
+
+
+
+**** Creating a blast database ****
+
+
+Now you can create your own BLAST databases
+You can create a BLAST database from any FASTA file
+
+There's some additional information in the README_BLAST file
+
+There are preformatted ones for things like nr on NCBI
+ftp://ftp.ncbi.nih.gov/blast/db/
+
+and you can download any sequenced microbial genome
+ftp://ftp.ncbi.nlm.nih.gov/genomes/Bacteria/
+
+If the databases are really big, like they are for nr, it's not something you'll want to do on your local computer.  A server or the HPCC is better for that.
+
+
+----  To create your own database from your own FASTA file ----
+
+Copy the file into the blast database directory you just referenced in the .ncbirc
+
+Here we'll use a test FASTA file (provided)
+
+test_fasta.fasta
+
+The 'formatdb' command creates the databases
+
+-i  is the input file
+-n is the name that you want for your database
+-p is the type of file (protein T, or nucleotide F)
+
+If you type 
+
+formatdb - 
+you get all the options
+
+Run the formatdb command with both -p T and -p F so you get both the nucleotide and protein database.  The different blast programs require the different databases.
+
+formatdb -i test_fasta.fasta -n test_fasta -p T
+
+and
+
+formatdb -i test_fasta.fasta -n test_fasta -p F
+
+Now if you look in that directory you have new files and those are your blast databases
+
+
+**** Running BLAST ****
+
+The command to run BLAST is 'blastall'
+
+If you type 
+blastall - 
+You get all the options
+
+This is a standard blastall command
+
+blastall -e 1e-05 -p tblastx -d test_fasta -i seqs.fa  -o seqs.blast
+
+-e is the e-value cutoff you want to use.  Any matches higher than that will not be returned
+-p is the program - tblastx, blastx, blastn or tblastn
+-d is the database
+-i is the input file
+-o is the output file
+-m is the output type you want
+   If you're parsing the output, then you want to use -m 8.  It outputs a tab delimited format that's easy to look through
+   The default shows you all the alignments
+
+If you do use -m 8 this is the information in each column
+
+# Query id # Subject id # % identity # alignment length # number of mismatches # number of gap openings # position of query start # position of query end # position of subject start # position of subject end # e-value of a hit # bit score of a hit  
+
+
+That's it, now you have your blast information and you can parse the BLAST output

SSH/index.rst

@@ -1,8 +1,11 @@
-SSH
+Useful UNIX tools
 =========
 
 **Updated and presented by: Tracy Teal**
 
+SSH
+--------
+
 
 What is ssh?
 ------------------
@@ -66,3 +69,39 @@ Now let's try a test example of logging on to a remote machine, running screen,
 running a blast at the command line, detaching from the screen and then coming 
 back to check on the results.
 
+Follow the instructions above for loggin in and starting screen
+
+Now to run command line blast
+
+You can download command line blast from NCBI.  It's called 'blastall'
+
+If you type 
+blastall - 
+You get all the options
+
+This is a standard blastall command
+
+blastall -e 1e-05 -p tblastx -d test_fasta -i seqs.fa  -o seqs.blast
+
+-e is the e-value cutoff you want to use.  Any matches higher than that will not be returned
+-p is the program - tblastx, blastx, blastn or tblastn
+-d is the database
+-i is the input file
+-o is the output file
+-m is the output type you want
+   If you're parsing the output, then you want to use -m 8.  It outputs a tab delimited format that's easy to look through
+   The default shows you all the alignments
+
+If you do use -m 8 this is the information in each column
+
+# Query id # Subject id # % identity # alignment length # number of mismatches # number of gap openings # position of query start # position of query end # position of subject start # position of subject end # e-value of a hit # bit score of a hit  
+
+So, we'll run blastall on a test dataset
+
+There are some test fasta files in the 'fasta' directory
+
+Go in to that directory.  You can use any of these files as the input fasta file.
+
+The blast database is: /home/swc/blastdb/nirK_ref_Rh
+
+blastall -e 1e-05 -p tblastx -d /home/swc/blastdb/nirK_ref_Rh -i oneseq.fasta  -o seqs.blast