Protein Sequence Analysis

The analysis of protein sequences provides the information about the preference of amino acid residues and their distribution along the sequences for understanding the secondary and tertiary structures of proteins and their functions. The identification of similar motifs in protein sequences would help to predict the structurally or functionally important regions. The profiles obtained with the single amino acid properties based on amino acid sequence would reveal the clustering of amino acids with similar property. Furthermore, the comparison of different amino acid sequences using alignment methods would enhance our knowledge about the availability of similar sequences, and these sequences could be used as a template for protein three-dimensional structure prediction.

2.1 Sequence alignment

The comparison of two proteins is mainly carried out by aligning the sequences or structures. In this method, a one-to-one correspondence is set up between the residues of the two proteins. The simplest observation is the global alignment of two sequences, in which the two proteins have maintained a correspondence over the entire length. An alternative is the local alignment in which the alignment is made only with the most similar part of the proteins.

An alignment of two sequences A and B must obey the following conditions: (i) All residues should be used in the alignment and all should be in the same order, (ii) align one residue from A with another from B, (iii) a residue can be aligned with a blank (-), and (iv) two blanks cannot be aligned. The different ways of aligning two sequences, VEITGEIST and PRETERIT, are shown in Figure 2.1. From these alignments, one could estimate the score for each aligned positions and hence the total score. The scoring scheme will be as follows: (i) Score = 1, if both the residues in the same positions of the sequences A and B are the same (e.g., in Alignment 1 [Figure 2.1], both the sequences A and B at position 3 are E, and hence it will have the score of 1), (ii) if the residues are different, score = 0 (e.g., position 1, the residues are V and P, respectively in sequences A and B), and (iii) score = − 1 if there is a gap in the alignment (e.g., positions 2 and 4 in Alignment 1). The added score for all the residues gives the net score for the aligned sequences. In alignments, the positioning of residues with similar properties (e.g., Val and Ile are hydrophobic, Glu and Asp are negatively charged, etc.) is used to find similar sequences (Eidhammer et al. 2004).

2.2 Programs for aligning sequences

Several computer programs have been developed for estimating the similarity score of two sequences and for finding similar sequences from available databases using pairwise and multiple alignments.

2.2.1 Basic Local Alignment Search Tool (BLAST)

Altschul et al. (1990) developed an approach for a rapid sequence comparison, basic local alignment search tool (BLAST), which directly approximates alignments that optimize a measure of local similarity and the maximal segment pair score. This algorithm has been applied in a variety of contexts, including straightforward DNA and protein sequence database searches, motif searches, gene identification searches, and in the analysis of multiple regions of similarity in long DNA sequences. In this method, the query protein sequence can be searched with several databases, including the nonredundant structures available in PDB, protein sequences at SWISS-PROT, etc. Furthermore, BLAST has several features such as (i) identifying protein sequences similar to the query, (ii) finding members of a protein family or building a custom position-specific scoring matrix, (iii) finding proteins similar to the query around a given pattern, (iv) finding conserved domains in the query, and (v) searching for peptide motifs. BLAST is available at http://www.ncbi.nlm.nih.gov/BLAST/. An example to identify protein sequences similar to the query is shown in Figure 2.2. BLAST has several options for querying a sequence:

(i) Accepts the sequence with accession number, gi, and FASTA format. The input data can be given by copying and pasting the details directly on the Web or by uploading a file from a local computer. Accession number is the number allotted in UniProt for each sequence (e.g., P61626); gi is a bar-separated NCBI sequence identifier (e.g., gi|48428995). A sequence in FASTA format begins with a single-line description, followed by lines of sequence data. The description line is distinguished from the sequence data by a greater than (“>”) symbol at the beginning. An example sequence in FASTA format is given below:
> gi|48428995|sp|P61626.1|LYSC_HUMAN RecName: Full=Lysozyme C MKALIVLGLVLLSVTVQGKVFERCELARTLKRLGMDGYRGISLANWMCLAKWESGYNTRATNYNAGDRSTDYGIFQINSRYWCNDGKTPGAVNACHLSCSALLQDNIADAVACAKRVVRDPQGIRAWVAWRNRCQNRDVRQYVQGCGV
The complete amino acid sequence in FASTA format has been provided in Figure 2.2a. It is also possible to specify a fragment of the sequence by providing a sub-range of the query sequence.

(ii) Allows selecting from a database to search against the input sequence. The nonredundant protein sequences (nr) have been selected as the database in Figure 2.2a.

(iii) The algorithm of the program can be selected and, for finding similar sequences, BLASTP is used.

(iv) It is possible to adjust several parameters: (a) displaying the maximum number of aligned sequences, expect threshold, and word size. Expect threshold (e-value) is the expected number of chance matches in a random model, and it is set at 10 as the default value. Word size is the length of the seed that initiates an alignment. In addition, scoring parameters can be selected for matrix, gap cost, and compositional adjustments. The substitution matrix is a key element in evaluating the quality of a pairwise sequence alignment, which assigns a score for aligning any possible pair of residues. Generally BLOSUM62 is used as the substitution matrix, which is a 20 × 20 matrix obtained for all possible substitutions of 20 amino acid residues (Table 2.1). It is based on a likelihood method by estimating the occurrence of each possible pairwise substitution using the biochemical character of amino acid residues (aliphatic, aromatic, positive charged, negative charged, polar, sulfur containing, etc., see Figure 1.2), and the development of BLOSUM62 has been described in Eddy (2004). The gap cost is a cost to create and extend a gap in an alignment. Furthermore, options are available to filter the low-complexity regions and mask query and lowercase letters in the sequence.