Oct. 6, 2014

FASTA User's Guide: $doc/fasta/fasta_guide.pdf
FASTA Manual $doc/fasta/fasta36.1.html
Bill Pearson's FASTA Web site: [http://www.people.Virginia.EDU/~wrp/pearson.html]

Example: Antifreeze Proteins

This tutorial assumes that local copies of the NCBI RefSeq RNA, RefSeq protein, Non-redundant nudcleotide (nt) and Non-redundant protein (nr) databases.

1. Read in the test sequence and generate the corresponding amino acid sequence

This tutorial will introduce the several types of sequence database searches, using the antifreeze protein from the sea raven, Hemitripterus americanus. The cDNA sequence for this gene is found in SRAAFP.gen.

Create a directory called 'db' and save this sequence to the db directory.

To read this sequence into bldna, choose 'File --> Open', and select SRAAFP.gen.

Since this tutorial will demonstrate database searches using both DNA and protein sequences as query sequences, we need to generate the amino acid sequence. This is done in two steps. First, we need to extract the protein coding sequence from the cDNA. (Remember that mRNAs contain both 5' and 3' untranslated sequences. If you look at SRAAFP.gen, you'll see that the CDS (protein coding sequence) runs from positions 10 to 597. We will use the FEATURES program to extract the CDS, and then translate the CDS using RIBOSOME.

Make sure SRAAFP is selected, and choose Database --> FEATURES (Extract by feature keys).

Click on 'Run' with the above settings. The CDS will appear in a new bldna window.

Note that the CDS begins with a start codon, indicating that this cDNA contains the entire protein coding region.

Next translate the CDS into protein. Select SRAAFP:CDS1, and choose DNARNA --> Ribosome.  By default, Ribosome sends the output to a new blprotein window.

2. The Database Menu



3. DNA vs. DNA

Select the DNA sequence SRAAFP:CDS1, and choose Database --> FASTA (DNA/RNA vs. DNA database).

First, choose a database. The quickest one to search is probably RefSeq RNA, which contains the transcripts and coding regions from most known genes encoding RNAs and proteins. Non-coding genomic regions, which make up the vast majority of eukaryotic genomes, are omitted from this dataset, which is consequently a lot smaller and faster to search.

Hint: For more distantly-related sequences, use K-tuple = 4 or 5.
The default K-tuple value of 6 ensures a faster search. However, some sensitivity is lost by using a larger K-tuple value. If you are concerned about finding sequences that are more distantly-related to your query, use K-tuple 4 or 5.

Note: The FASTA, FASTX/Y, and TFASTA search BOTH strands of DNA database sequences, by default.

Output will go to two windows. The FASTA report will appear in a text editor,

and the GenBank accession numbers of database hits will appear in blnfetch, a BioLegato interface specialized for retrieving database entries.

The results in both windows are in temporary files that will be deleted once each window is closed. To save the FASTA report from the text editor, choose File --> Save As, and type in a filename. A good name that keeps track of the sequence used, the program run, and the type of file would be SRAAFP.RefSeqRNA.fasta.

Similarly, the list of accession numbers can be saved from blnfetch by choosing File --> Save Selection As SRAAFP.RefSeqRNA.fasta.acc.

Note that the file format was set to TSV, but the file extension is .acc. Since the file only contains a single column, the distinction between a TSV file and a text file is moot.

A quick word on CSV, TSV and text files, and BioLegato
Most spreadsheet programs can read and write formats called CSV and TSV. CSV files are comma-separated value text files, in which each line in the file represents a row in a spreadsheet. The values for each column in a row are separated by commas. TSV files are TAB-separated value text files. In these files, values in each row are separated by TAB characters. TSV files are probably safer to use, because CSV files make it impossible to include commas within fields.

BioLegato programs such as blnfetch, blpfetch and blncbi have a table canvas, that works like a spreadsheet. In these programs, the Open menu can only read text files with a .csv or .tsv file extension. To read files with other extensions eg. .acc, you need to use File --> Import Data from CSV file. Text files such as the one generated above will be read using this approach.

Retrieving your hits
If you decide to retrieve all the hits, this can be done from blnfetch. Choose Edit --> Select All, and then Database --> SeqFetch. By default, SEQFETCH is set to retrieve GenBank entries to a text editor. If you had a large number of entries, it would probably be best to sent output to a file.

Entries retrieved to a text editor must be saved if you wish to keep them. An example can be found in SRAAFP.RefSeqRNA.fasta.gen.

4. Protein vs. Protein

DNA vs. DNA searches are far less sensitive than protein vs. protein searches. DNA searches only work for closely-related sequences.

Consequently, protein vs. protein searches are likely to detect more distantly-related genes. As well, since amino acid sequences are only 1/3 the length of the corresponding DNA sequences, protein searches require only (1/3)2 = 1/9 the computational time.

In your blprotein window, select the amino acid sequence SRAAFP:CDS1, and choose Database --> FASTA (protein vs. protein database).

Note that by default, accession numbers will be sent to blpfetch, a BioLegato interface for retrieving protein entries, while the FASTA report will go to a text editor.

For DNA searches a very simple scoring scheme is used, in which +1 is added to the score for a match, and -1 is added to the score for a mismatch. The greater diversity of amino acids lends itself to more complex scoring schemes. The menu shows a variety of protein scoring matrices. The Blosum matrices were constructed using distantly-related sequences. If you need a highly sensitive search, use a low-numbered Blosum matrix. The PAM matrices were constructed from a set of closely-related sequences. For a high-sensitivity search, use PAM250, or PAM250 Gonnet (based on a more recent dataset.). For closely-related sequences, use PAM120.

More information of scoring matrices can be found in
Hugh B. Nicholas Jr., David W. Deerfield II., and Alexander J. Ropelewski, (1998) A Tutorial on Searching Sequence Databases and Sequence Scoring Methods Developed by the Biomedical Supercomputing Initiative of the Pittsburgh Supercomputing Center.[].

The output from this run can be found in SRAAFP.RefSeqPro.fasta and SRAAFP.RefSeqPro.fasta.acc.
Hint: Oligopeptides need logarithmic score weighting
For very short query sequences such as oligopeptides, unweighted scores would not exceed the minimum cutoff values needed to appear in the output. The FASTA programs allow you to specify a logarithmic weighting ratio that gives short query sequences higher matches. The score is weighted by the natural log of the query length divided by the natural log of the database size.

5. Protein vs. Translated DNA

TFASTA is the most sensitive method for searching DNA sequence databases. Each DNA sequence is translated on the fly in either 3 or 6 reading frames (ie. one or both strands) . ALL of the sequence is translated, whether coding or non-coding, intron or exon.

By default, TFASTA is set to search the GenBank Non-redundant nucleotide database.


Two versions of TFASTA are available.  TFASTX allows 3-base insertions (ie. 1 codon), while TFASTY allows 1 or 2 base insertions. TFASTY is therefore the most sensitive program, but it is also slower. TFASTY would be especially good at detecting similarities between a query protein and ESTs, since ESTs typically have more frameshifts than most sequences.

Note on automatic translation:

Programs that translate sequences on the fly (eg. TFASTX, TBLASTN, TBLASTX) have no knowledge whatsoever about gene structure (ie. exons, introns, 5' UTR, 3'UTR).  All these programs do is take every group of 3 nucleotides and assign a codon to it. Even stop codons are represented by an asterisk (*).  Consequently, non-coding sequences and non-coding open reading frames are translated into meaningless amino acid sequences.

Note: This search took 19 min. on a 16-core  AMD Opteron Linux system with 64 Gb of memory, with many sessions running by other users. Be prepared for a bit of a wait. This is primarily due to the size of the nt database. By comparison, the same search run on a weekday using blastn at NCBI took about 3 min. to return results. These results will vary greatly, depending on the user load on each system at a given time.

Hint: Send output from long-running searches to a file.

bldna runs all database searches in the background, ie. as independent jobs. If you quit bldna or blprotein or even logout, the job will run to completion. For long-running searches, it is usually best to send output directly to a file. Simply select 'Output file', and type an Output file name. In the example above, output will be directed to two files: SRAAFP.nt.tfastax, which contains the FASTA report, and, which contain the GI numbers of the hits. Notice that biolegato automatically appends the .nam file extension.

6. DNA vs. Protein

FASTX and FASTY are the converse of TFASTX and TFASTY, in that they translate a DNA query sequence, allowing either codon-sized gaps (TFASTX) or 1 or 2 base gaps (TFASTY). TFASTY is therefore the slower, but more sensitive. These programs are particularly well-suited for comparing a sequence that may have frameshifts (eg. an EST) with a protein database. To speed things up in this demonstration, we'll search the RefSeq Protein database.

Again, since this search will take more time because of fasty, output goes to and

7. Retrieving Hits

1. blnfetch, blpfetch - retrieve sequences from NCBI using NCBI Eutils

When you have saved the names or accession numbers of database hits, the sequences can easily be retrieved using blnfetch for nucleotide sequences, and blpfetch for protein sequences.

For example, to retrieve protein sequence hits from the fasty search in above, start blpfetch at the command line by typing blpfetch.

Read in the names from using File --> Import Data from CSV file.    

Since this is a big list containing 2598 accession numbers, we'll limit the retrieval to the top 20 hits, and send the output to blprotein. Select the first 20 accession numbers by clicking on the cell in the first row, holding down the SHIFT key, and then clicking on the 20 th accession number.

Next, choose Database --> SeqFetch. We'll retrieve output in GenBank format, but sent to blprotein so that we can immediately begin working with the sequences.

If we had wanted to retrieve the entire list of sequences, it would have been best to send output to a file.

The exact same procedure can be used to retrieve DNA sequences using a list of accession or GI numbers as input for blnfetch.

2. NCBI Batch Entrez - retrieve sequences from NCBI
web site

The NCBI web site offers another way to retrieve the GenBank entries associated with your accession numbers, we'll use Batch Entrez at As an example, we'll use
SRAAFP.RefSeqRNA.fasta.acc, the file of hits from searching the SRAAFP protein against the RefSeq RNA division of GenBank using fasta. (You may need to save this file in the current working directory.) Go to the batchentrez link above and click on Browse. In the file chooser, select SRAAFP.RefSeqRNA.fasta.acc.

Now, click on Retrieve. A message will appear saying something like "Retrieve records for 20 UID(s). Click on that link.

Click on Send to and fill in the following:

Destination: File
Format: GenBank

When you click on Create File, a file chooser will pop up.

The default name will be something like "", but it is good practice to give it a name consistent with the rest of the files in this exercise. A good name would be


NOTE: Remember to save this file in the current working directory! Your file chooser will show a different location from the one shown a right.

This file will contain full GenBank flatfile entries for all of the fasta hits.