BLAST. Jon-Michael Deldin. Dept. of Computer Science University of Montana Mon

Transcription

1 BLAST Jon-Michael Deldin Dept. of Computer Science University of Montana Mon Jon-Michael Deldin (UM) BLAST Mon 1 / 23

2 Outline 1 Goals 2 Setting up your project 3 Running BLAST 4 Planning 5 Process Jon-Michael Deldin (UM) BLAST Mon 2 / 23

3 Topic 1 Goals 2 Setting up your project 3 Running BLAST 4 Planning 5 Process Jon-Michael Deldin (UM) BLAST Mon 3 / 23

4 Goals get familiar with BLAST Jon-Michael Deldin (UM) BLAST Mon 4 / 23

5 Goals get familiar with BLAST show how you can integrate scripting and external tools to save time Jon-Michael Deldin (UM) BLAST Mon 4 / 23

6 Topic 1 Goals 2 Setting up your project 3 Running BLAST 4 Planning 5 Process Jon-Michael Deldin (UM) BLAST Mon 5 / 23

7 The easy way If you re on a Mac or Linux machine, run setup.pl from your project root to create data directory Be sure to read through the script before executing it. Jon-Michael Deldin (UM) BLAST Mon 6 / 23

8 The easy way If you re on a Mac or Linux machine, run setup.pl from your project root to create data directory download all sequences Be sure to read through the script before executing it. Jon-Michael Deldin (UM) BLAST Mon 6 / 23

9 The easy way If you re on a Mac or Linux machine, run setup.pl from your project root to create data directory download all sequences download the query file Be sure to read through the script before executing it. Jon-Michael Deldin (UM) BLAST Mon 6 / 23

10 The easy way If you re on a Mac or Linux machine, run setup.pl from your project root to create data directory download all sequences download the query file create the database Be sure to read through the script before executing it. Jon-Michael Deldin (UM) BLAST Mon 6 / 23

11 The easy way If you re on a Mac or Linux machine, run setup.pl from your project root to create data directory download all sequences download the query file create the database index the database Be sure to read through the script before executing it. Jon-Michael Deldin (UM) BLAST Mon 6 / 23

12 The easy way If you re on a Mac or Linux machine, run setup.pl from your project root to create data directory download all sequences download the query file create the database index the database create stub files Be sure to read through the script before executing it. Jon-Michael Deldin (UM) BLAST Mon 6 / 23

13 Installing BLAST Download one of the installers from the NCBI (links on the HTML version) Jon-Michael Deldin (UM) BLAST Mon 7 / 23

14 Project layout Example (Tree) README data/ NC_ fna NC_ fna NC_ fna NC_ fna anthraxdb.fna querynuc.txt setup.pl src/ query_nuc.pl query_prot.pl Jon-Michael Deldin (UM) BLAST Mon 8 / 23

15 Project layout Example (Tree) README data/ NC_ fna NC_ fna NC_ fna NC_ fna anthraxdb.fna querynuc.txt setup.pl src/ query_nuc.pl query_prot.pl Explanation File README data/ setup.pl src/ src/query_nuc.pl src/query_prot.pl Purpose How to run your program, etc. Contains data for your project Automate setting up data Source code Performs nucleotide searches Performs protein searches Jon-Michael Deldin (UM) BLAST Mon 8 / 23

16 Downloading sequences Jon-Michael Deldin (UM) BLAST Mon 9 / 23

17 Downloading sequences Jon-Michael Deldin (UM) BLAST Mon 9 / 23

18 Downloading sequences Downloading from the command-line Jon-Michael Deldin (UM) BLAST Mon 9 / 23

19 Downloading sequences Downloading from the command-line OS X Use curl -O URL Jon-Michael Deldin (UM) BLAST Mon 9 / 23

20 Downloading sequences Downloading from the command-line OS X Use curl -O URL Linux Use wget URL Jon-Michael Deldin (UM) BLAST Mon 9 / 23

21 Downloading sequences Downloading from the command-line OS X Use curl -O URL Linux Use wget URL Windows Use your browser Jon-Michael Deldin (UM) BLAST Mon 9 / 23

22 Downloading sequences Downloading from the command-line OS X Use curl -O URL Linux Use wget URL Windows Use your browser Jon-Michael Deldin (UM) BLAST Mon 9 / 23

23 Downloading sequences Downloading from the command-line OS X Use curl -O URL Linux Use wget URL Windows Use your browser Files Jon-Michael Deldin (UM) BLAST Mon 9 / 23

24 Downloading sequences Downloading from the command-line OS X Use curl -O URL Linux Use wget URL Windows Use your browser Files NC_ fna Jon-Michael Deldin (UM) BLAST Mon 9 / 23

25 Downloading sequences Downloading from the command-line OS X Use curl -O URL Linux Use wget URL Windows Use your browser Files NC_ fna NC_ fna Jon-Michael Deldin (UM) BLAST Mon 9 / 23

26 Downloading sequences Downloading from the command-line OS X Use curl -O URL Linux Use wget URL Windows Use your browser Files NC_ fna NC_ fna NC_ fna Jon-Michael Deldin (UM) BLAST Mon 9 / 23

27 Downloading sequences Downloading from the command-line OS X Use curl -O URL Linux Use wget URL Windows Use your browser Files NC_ fna NC_ fna NC_ fna NC_ fna Jon-Michael Deldin (UM) BLAST Mon 9 / 23

28 Preparing the DB Need to merge all of our FNA files into anthraxdb.fna so BLAST can search it Mac & Linux cd data cat *.fna > anthraxdb.fna Windows cd data copy /a *.fna anthraxdb.fna Jon-Michael Deldin (UM) BLAST Mon 10 / 23

29 Index the database From your project s working directory (i.e., above data), type makeblastdb -in data/anthraxdb.fna -dbtype nucl Jon-Michael Deldin (UM) BLAST Mon 11 / 23

30 Download the query file Save querynuc.txt to data/ (we ll use it in the next tutorial) Jon-Michael Deldin (UM) BLAST Mon 12 / 23

31 Topic 1 Goals 2 Setting up your project 3 Running BLAST 4 Planning 5 Process Jon-Michael Deldin (UM) BLAST Mon 13 / 23

32 From the shell Do a search against the database: blastn -db data/anthraxdb.fna -query data/query.txt look at the E-values (smaller is better) Jon-Michael Deldin (UM) BLAST Mon 14 / 23

33 Topic 1 Goals 2 Setting up your project 3 Running BLAST 4 Planning 5 Process Jon-Michael Deldin (UM) BLAST Mon 15 / 23

34 Remember our goal We are trying to automate queries against BLAST to determine whether 100 fragments are from the A0248 strain. Jon-Michael Deldin (UM) BLAST Mon 16 / 23

35 Example Run $ perl query_nuc.pl best hit: ambiguous best hit: ambiguous : best hit: Bacillus anthracis str. A0248 : : best hit: ambiguous votes for ambiguous: XX votes for Bacillus anthracis str. A0248: YY Jon-Michael Deldin (UM) BLAST Mon 17 / 23

36 Input & Output input results from a BLAST command Jon-Michael Deldin (UM) BLAST Mon 18 / 23

37 Input & Output input results from a BLAST command output how many hits were ambiguous or conclusive Jon-Michael Deldin (UM) BLAST Mon 18 / 23

38 Process (pseudocode) 1 Read the sequences in from the query file (querynuc.txt) Jon-Michael Deldin (UM) BLAST Mon 19 / 23

39 Process (pseudocode) 1 Read the sequences in from the query file (querynuc.txt) 2 For each query sequence: Jon-Michael Deldin (UM) BLAST Mon 19 / 23

40 Process (pseudocode) 1 Read the sequences in from the query file (querynuc.txt) 2 For each query sequence: 1 Write the query to a text file to use in a BLAST command Jon-Michael Deldin (UM) BLAST Mon 19 / 23

41 Process (pseudocode) 1 Read the sequences in from the query file (querynuc.txt) 2 For each query sequence: 1 Write the query to a text file to use in a BLAST command 2 Run the BLAST command Jon-Michael Deldin (UM) BLAST Mon 19 / 23

42 Process (pseudocode) 1 Read the sequences in from the query file (querynuc.txt) 2 For each query sequence: 1 Write the query to a text file to use in a BLAST command 2 Run the BLAST command 3 Parse the output to determine what strain Jon-Michael Deldin (UM) BLAST Mon 19 / 23

43 Process (pseudocode) 1 Read the sequences in from the query file (querynuc.txt) 2 For each query sequence: 1 Write the query to a text file to use in a BLAST command 2 Run the BLAST command 3 Parse the output to determine what strain 3 Print out how many ambiguous and conclusive strains were found Jon-Michael Deldin (UM) BLAST Mon 19 / 23

44 Topic 1 Goals 2 Setting up your project 3 Running BLAST 4 Planning 5 Process Jon-Michael Deldin (UM) BLAST Mon 20 / 23

45 At the top... use strict; use warnings; Jon-Michael Deldin (UM) BLAST Mon 21 / 23

46 At the top... use strict; use warnings; my $db = data/anthraxdb.fna ; # path to BLAST DB Jon-Michael Deldin (UM) BLAST Mon 21 / 23

47 At the top... use strict; use warnings; my $db = data/anthraxdb.fna ; # path to BLAST DB my $query_fn = tmp_query.txt ; # file we generate each seq Jon-Michael Deldin (UM) BLAST Mon 21 / 23

48 At the top... use strict; use warnings; my $db = data/anthraxdb.fna ; # path to BLAST DB my $query_fn = tmp_query.txt ; # file we generate each seq # this is defined in the subroutines tutorial -- paste # the definition into your file = fasta_to_array( data/querynuc.txt ); Jon-Michael Deldin (UM) BLAST Mon 21 / 23

49 At the top... use strict; use warnings; my $db = data/anthraxdb.fna ; # path to BLAST DB my $query_fn = tmp_query.txt ; # file we generate each seq # this is defined in the subroutines tutorial -- paste # the definition into your file = fasta_to_array( data/querynuc.txt ); my $num_ambiguous = 0; # number of ambiguous hits my $num_conclusive = 0; # number of conclusive hits Jon-Michael Deldin (UM) BLAST Mon 21 / 23

50 At the top... use strict; use warnings; my $db = data/anthraxdb.fna ; # path to BLAST DB my $query_fn = tmp_query.txt ; # file we generate each seq # this is defined in the subroutines tutorial -- paste # the definition into your file = fasta_to_array( data/querynuc.txt ); my $num_ambiguous = 0; # number of ambiguous hits my $num_conclusive = 0; # number of conclusive hits # command to r un for each query (i.e., 100 times) my $cmd = "blastn -db $db -query $query_fn -evalue 1e-10"; Jon-Michael Deldin (UM) BLAST Mon 21 / 23

51 In the middle... for my $query { # create a BLAST-query for the current sequence ($query) #...write it to $query_fn # execute BLAST # my $result = $cmd ; } # see what we got (parse the output) Jon-Michael Deldin (UM) BLAST Mon 22 / 23

52 At the bottom... print "Total ambiguous: $num_ambiguous\n"; print "Total conclusive: $num_conclusive\n"; Jon-Michael Deldin (UM) BLAST Mon 23 / 23