Direct Comparison of two FASTA sequences using bl2seq

Direct Comparison of Two FASTA Sequences Using bl2seq

Tao Tao, PhD
User Services, NCBI, NLM, NIH

TOC

1. Introduction
2. Installation and setup
3. Common usage examples
4. Feedback
5. Appendix: program parameters of bl2seq

The program bl2seq is designed to compare two input FASTA sequences directly bypassing the formatdb step. This allows users to quickly assess the similarities between two input sequences. The program is capable of performing six types of blast searches: blastn, blastp, blastx, tblastn, tblastx, and megabalst in place of blastn.

The input sequences can be provided as the FASTA sequence files name or in the accession.version format under '-A T' setting, if the client server function can connect to NCBI.

2. Installation and setup

bl2seq does not require program specific setup if the client server function is not used. For users behind firewalls, using the client server function will require addtional setup. Specifically, the following lines need to be added to .ncbirc or ncbi.ini.

[CONN]
FIREWALL=TRUE

[NET_SERV]
SRV_CONN_MODE=SERVICE

This also requires that the corresponding ports be open for connection to certain NCBI IP addresses. Refer to Searching Against NCBI Databases With blastcl3 for more information on firewall and .ncbirc (ncbi.ini) configuration.

3. Common usage examples

bl2seq is a convenient tool for directly comparison of two local sequences in all possible combination of sequence types. Here we will discuss a few common comparisons using this tool.

One search bl2seq can perform is the direct comparison of two nucleotide sequences to find out the differences present between teh two sequences. The two sequences can be a mutated clone and its parent, a field isolate and the reference strain, or a the mRNA and its genomic counterpart.

bl2seq -i mutant.seq -j parent.seq -p blastn -F F -o mut_parent.out

We can also use bl2seq directly compare PCR primers against their target sequence to identify the annealing position and amplicon size. To search two primers together, we need to convert them to the following format. The reverse primer does not have to be converted to its reverse complement since bl2seq does this automatically.

>Human MLH1 primer pair 
TGCACTGTGGGATGTGTTCT
NNNNNNNNNNNNNNNNNNNN
AATCAATCCACTGTGTATAAAGGAA

If the above primer pair is saved in MLH_primer.seq, and the target sequence is in MLH.seq, we can directly compare the two using the following command line. The result is saved in amplicon. Since the refseq mRNA for human MLH1 is NM_000249, we can use the second command line if the client server connection is set up correctly.

bl2seq -i MLH_primer.seq -j MLH.seq -p blastn -F F -o amplicon.out

bl2seq -i MLH_primer.seq -j NM_000249.2 -p blastn -F F -o amplicon.out -A T

If both primers map to the target well, then the amplicon can be calculated using the largest and smallest coordinates as given in the alignment output as shown below.

Reverse Primer Match

Forward Primer Match

Score = 50.1 bits (25), Expect = 1e-010
 Identities = 25/25 (100%)
 Strand = Plus / Minus
 
Query: 51   aatcaatccactgtgtataaaggaa 75
            |||||||||||||||||||||||||
Sbjct: 2496 aatcaatccactgtgtataaaggaa 2472

Score = 40.1 bits (20), Expect = 1e-007
 Identities = 20/20 (100%)
 Strand = Plus / Plus
 
Query: 1    tgcactgtgggatgtgttct 20
            ||||||||||||||||||||
Sbjct: 2345 tgcactgtgggatgtgttct 2364

Amplicon size is the difference of the two underlined coordinates plus 1, i.e. (2496 � 2345) + 1 = 152 bp

We can also use bl2seq to identify the region in a BAC clone to which a given mRNA sequence matches.

bl2seq -p blastn -i my_mRNA -j my_BAC -F "m L" -m T -o mRNA_BAC.out

bl2seq -p blastn -i NM_000249.2 -j AC006583.31 -F "m L" -m T -o hs_MLH1.out

In the first command line above, we use megablast and low complexity filter to reduce spurious hits. For clear result, we also add "m" to the filter string to allow extension through the filter masked region. In the second command line, we use the client server function to map retrieve the human MLH1 mRNA and a BAC clone containing this gene before matching them. For precise mapping of the exon/intron junctions, however, we should use splign or spidey, two specialized tools designed for this purpose. More information is available at: spideydoc.html, and
splign document.

Sometimes two nucleotide sequences may not have similarity detectable by bl2seq's blastn function. However, if both of them encodes for protein products, we can use the tblastx program function to detect the potential similarity shared by these products.

For example, the following two sequences have no detectable nucleotide similarity when compared using bl2seq's blastn function.

>Hypera postica cysteine proteinase, partail sequence
GGGAGCCAAATTCCAGGCCTTCAAGTTGGAGCATGGCAAGACCTACTTAAACCAAGCTGAGGAAAGCAAG
CGCTTTAACATCTTTACTGACAACGTACGCGCTATTGAAGCACACAATGCCCTCTACGAGCAAGGAAAAG
TATCCTACAAAAAAGGTATCAATAAATTCACTGACATGTCTCAAGAAGAGTTCAAGACAATGCTCACTCT
CAGCGCATCTAGAAAACCAACTTTGGAAACTACTTCATACGTAAAAACCGGTGTTGAAATCCCATCATCT
GTTGACTGGAGAAAAGAAGGTCGAGTAACTGGAGTCAAAGATCAAGGCGATTGTGGATCATGCTGGGCAT
TCTCTATCACTGGATCAACCGAAGGCGCCTACGCCCGTAAATCTGGGAAACTTGTTTCTCTTTCTGAACA
ACAATTGATAGACTGCTGCACTGATACAAGTGCAGGATGTGATGGTGGATCACTAGACGACAATTTTAAA
TACGTCATGAAGGATGGTCTTCAGTCTGAAGAAAGCTACACCTACAAGGGTGAGGATGGAGCATGCAAAT
ACAACGTTGCAAGTGTTGTAACTAAAGTCAGCAAATACACTTCCATTCCAGCAGAAGACGAAGATGCTCT
TCTTGAGGCTGTAGCTACTGTAGGACCAGTATCTGTTGGCATGGATGCTAGCTACCT
>Boophilus microplus cathepsin L-like proteinase, partial sequences
ACTGTTGCTGCAAGCTCTCAAGAAATCCTACGCACCCAATGGGAGGCATTTAAAACTACCCACAAAAAAT
CCTACCAGTCACACATGGAGGAGCTCCTGAGGTTCAAGATTTTCACGGAGAACAGCCTAATCATTGCCAA
GCACAACGCTAAGTACGCCAAGGGTCTCGTTTCTTACAAGCTCGGAATGAACCAGTTCGGCGATCTGCTG
GCACACGAATTTGCCAGGATCTTCAACGGTCACCACGGAACCCGCAAAACCGGTGGATCGACCTTCCTGC
CACCAGCAAACGTCAATGACAGCAGCCTGCCAAAAGTTGTCGACTGGCGCAAAAAAGGAGCTGTCACACC
TGTCAAGGACCAGGGACAGTGCGGGTCTTGCTGGGCCTTCAGTGCAACTGGATCTCTGGAGGGACAGCAT
TTTCTGAAGAACGGGGAGCTCGTTTCACTCAGTGAACAAAACTTGGTCGACTGTTCTCAGTCCTTCGGCA
ACAATGGTTGTGAAGGTGGTCTCATGGAGGACGCCTTTAAGTACATCAAGGCAAACGATGGTATCGACAC
GGAAAAAAGCTACCCATATGAGGCTGTGGATGGCGAGTGTCGTTTCAAGAAGGAAGATGTTGGAGCAACC
GACACCGGCTATGTGGAAATCAAGGCGGGTTCTGAGGTTGACCTGAAGAAGGCCGTCGCTACGGTCGGCC
CCATCTCTGTGGCTATTGACGCTAGTCACTCATCATTCCAGCTGTATTCCGAAGGAGTGTACGATGAGCC
CGAGTGTAGCAGTGAAGATCTGGACCACGGTGTACTCGTTGTCGG

However, the similarity at the protein product level is readily detectable, if we use the translated search function tblastx. The command line to use would be the following if we save them as seq_1.txt and seq_2.txt, respectively.

bl2seq -i seq_1.txt -k seq_2.txt -p tblastx -F F

A common misuse of bl2seq is to try to identify patterns or sequece motifs that might be present in the target sequence. This type of attempts often fails since bl2seq does not handle ambiguous code. A more suitable and yet little known tool from the standalone BLAST package is seedtop. Please refer to Pattern Search With seedtop for more information.

4. Feedback

For questions and comments on this document and BLAST in general, please send them to:

blast-help@ncbi.nlm.nih.gov

Questions and comments on other NCBI resources should be addressed to:

info@ncbi.nlm.nih.gov

5. Appendix: program parameters of blseq

We control behavior of a given BLAST program through command line parameter and value pairs. The convention is -A value, where the dash plus single letter marks the parameter with value following it. In this section, we will explain the available bl2seq parameters individually, one per table.

Table 5.1
Option	-i
Function	First Input query sequence
Default	-
Example	To search with my_primer as the first query, use: -i my_primer

Note
This parameter must be set. Since bl2seq treats the second input query as database, it is recommended that the short query be provided to this option. When -A T is used, -i takes accession.version, such as in "-i NM_000249.1".

Table 5.2
Option	-j
Function	Second Input query sequence
Default	-
Example	To search with my_target as the second query, use: -i my_target

Note
This parameter must be set. Large query should be provided to this parameter since bl2seq treats it as the database.

Table 5.3
Option	-p
Function	Program function name, must be set
Default	-
Example	To search with tblastn, use: -p tblastn

Note
Program and Query type combination

Program         blastn    blastp    blastx   tblastn   tblastx
Query (-i)        NT        AA        NT        AA        NT
2nd Query (-j)    NT        AA        AA        NT        NT

Table 5.4
Option	-g
Function	Perform gapped alignment
Default	T
Example	To disable gapping, use: -g F

Note
Default is to do gapped alignment.

Table 5.5
Option	-o
Function	To save alignment to specified output file
Default	stdout
Example	To save search result to my_output, use: -o my_output

Note
Default is to print the output to the screen, which can be redirected to file or piped to other downstream process.

Table 5.6
Option	-d
Function	Theoretical database size
Default	0
Example	To use a theoretical database size of 2000000, use: -d 2000000

Note
Default is to use actual size of the second query. We can use this parameter to provide the actual size of a real database such as protein nr to get a more realistic Expect value for the returned protein alignment.

Table 5.7
Option	-a
Function	SeqAnnot output file
Default	Optional
Example	To save the this to my_seqalign, use: -a my_seqalign

Note
The output is an ASN.1 file.

Table 5.8
Option	-G
Function	Cost ot open a gap
Default	Varies
Example	To increase the gap open penalty to -5, use: -G -5

Note
Defaults for various -p settings

Program    blastn    blastp    blastx    tblastn    tblastx    megablast
Value        5         11        11         11         11          0

Table 5.9
Option	-E
Function	Cost to extend a gap
Default	-1
Example	To increase the gap extension penalty to -2, use: -E 2

Note
Defaults for various -p settings

Program    blastn    blastp    blastx    tblastn    tblastx    megablast
Value        2         1         1         1           1          0

Table 5.10
Option	-X
Function	X dropoff value for gapped alignment (in bits)
Default	0
Example	To increase this X dropoff to 50, use: -X 50

Note
Zero invokes the following default

Program    blastn  megablast  tblastx     all others
Value        30      20          0           15

Table 5.11
Option	-W
Function	Word size
Default	0
Example	To decrease the nucleotide search word size to 8, use: -W 8

Note
Zero invokes the following defaults

Program    blastn   megablast   all others
Value        11        28           3

Table 5.12
Option	-M
Function	Score matrix to use
Default	BLOSUM62
Example	To use PAM30 for short peptide search, use: -M PAM30

Note
bl2seq only supports BLOSUM45, BLOSUM62, BLOSUM80, PAM30, and PAM70.

Table 5.13
Option	-q
Function	Penalty for a nucleotide mismatch
Default	-3
Example	To decrease this to -2, use: -q -2

Note
For blastn only.

Table 5.14
Option	-r
Function	Reward for a nucleotide match
Default	1
Example	To increase this to 2, use: -r 2

Note
For blastn only.

Table 5.15
Option	-F
Function	Filter query sequence
Default	T
Example	To turn off filter, use: -F F

Note
Accepted strings: T, F, D, L, R, V, S, C, and m. See Section 6.4 of the "BLAST URLAPI" document for details.

Table 5.16
Option	-e
Function	Expect value
Default	10.0
Example	To increase this for short primer search, use: -e 1000

Note
This controls the search stringency. To increase the stringency, reduce the value. To reduce the stringency, do the reverse.

Table 5.17
Option	-S
Function	Query strands to use in the search (1st query)
Default	3
Example	To search with only the input strand, use: -S 1

Note
For nucleotide search with blastn only: 3 is both, 1 is top, 2 is bottom.

Table 5.18
Option	-T
Function	Produce HTML output
Default	F
Example	To produce HTML output viewable through browser, use: -T T

Note
Hit to NCBI sequence will be hot linked to its Entrez record if -A T and accession.version are used.

Table 5.19
Option	-m
Function	Use megablast for search
Default	F
Example	To trigger megablast algorithm, use: -m T

Note
For blastn only. Search will be faster, but less sensitive.

Table 5.20
Option	-Y
Function	Effective length of the search space
Default	0
Example	To use an effective search space of 1000000, use: -Y 1000000

Note
The default is to use the actual search space defined by the input queries.

Table 5.21
Option	-t
Function	Length of the largest intron allowed in tblastn for linking HSPs
Default	0
Example	To link HSPs 2000 bps apart, use: -t 2000

Note
Zero disables HSP linking.

Table 5.22
Option	-I (upper case i)
Function	Location on first query
Default	Optional
Example	To search subsequence 100-200 of the first query, use: -I 100,200

Note
N/A.

Table 5.23
Option	-J
Function	Location on second query
Default	Optional
Example	To search subsequence 1000-2000 of the second query, use: -J 1000,2000

Note
This is the only program that allows users to specify a subsequence in the "database" entry.

Table 5.23
Option	-D
Function	Output format
Default	0
Example	To see the tabular output, use: -D 1

Note
An input of 0 generates pairwise display and an input of 1 generates tabular display. Currently bl2seq does not support XML.

Table 5.24
Option	-U
Function	Use lower case filtering for thequery sequences
Default	F
Example	To enable lowercase filtering, use: -U T

Note
This is for the first query. Make sure there is uppercased region available in that sequence.

Table 5.25
Option	-A
Function	Input sequences in the form of accession.version
Default	F
Example	To enable this parsing and retrieving, use: -A T

Note
The input to -i and/or -j should be accession.version, such as in -i NM_000249.2.

Table 5.26
Option	-V
Function	Force use of the legacy BLAST engine
Default	F
Example	To enable this, use: -V T

Note
Not recommended.