103. A protein sequence was isolated from a novel source. During an initial sequencesimilarity
search through BLAST, no homologous sequence was identified. A furtherBLAST search should
be run by changing the scoring matrix to:
(a) BLOSUM62,
(b) BLOSUM80,
(c) PAM250,
(d) PSSM

---

Introduction

When working with protein sequence data, one of the first steps in identifying potential homologs is performing a BLAST (Basic Local Alignment Search Tool) search. However, sometimes a search may yield no significant matches, even when you’re confident that the sequence should be similar to known proteins. This scenario often calls for tweaking the search parameters, specifically the scoring matrix. In this article, we’ll discuss the importance of choosing the right scoring matrix, particularly in situations where your initial BLAST search doesn’t return any homologous sequences.

---

Why Change the Scoring Matrix?

BLAST works by comparing the query sequence against a database of sequences using a scoring matrix. The scoring matrix determines how alignments between sequences are scored, taking into account substitutions (or differences) between amino acids. If the initial BLAST search doesn’t return any homologs, it might be because the default matrix is too stringent or not well-suited for the sequence in question. Adjusting the scoring matrix can make a significant difference in detecting homologous sequences that were previously missed.

Common Scoring Matrices in BLAST

There are several different scoring matrices that are commonly used in sequence similarity searches, and choosing the right one can depend on the nature of the sequences being compared:

• BLOSUM62: The BLOSUM (Blocks Substitution Matrix) family is one of the most widely used for protein sequence comparison. The BLOSUM62 matrix is particularly well-suited for sequences with moderate similarity (60-70% identity), making it a good default choice for general searches. If your sequence didn’t return matches using a default matrix, switching to BLOSUM62 might help in many cases.

• BLOSUM80: This matrix is more stringent and is better for comparing highly conserved sequences (greater than 80% identity). It’s less likely to detect distant homologs but is more sensitive when you’re working with closely related sequences. If your sequence is highly conserved, using BLOSUM80 can yield better results.

• PAM250: The PAM (Point Accepted Mutation) matrices are derived from the idea of evolutionary changes, and PAM250 is one of the matrices for sequences with high divergence (more than 85% difference). It can be more useful when comparing distantly related sequences. If no homologous sequences are found with more stringent matrices, trying PAM250 might help uncover more distant relationships.

• PSSM (Position-Specific Scoring Matrix): Unlike the others, PSSM is not a static matrix but is generated dynamically based on multiple sequence alignments for a specific protein family. It can be particularly useful when you’re searching for homologs in a very specific protein family or when the protein has some known conserved motifs. However, PSSMs are not always applicable to general BLAST searches and are typically used for more targeted research.

---

Best Matrix for a Sequence with No Homologs

In the case where a protein sequence from a novel source returns no homologous sequences in the initial search, it’s often best to switch to a more permissive scoring matrix like PAM250. The PAM250 matrix is designed for comparisons involving more divergent sequences, so it’s well-suited for identifying distant relatives that might not share obvious similarities with well-characterized proteins. By adjusting the scoring matrix to PAM250, you increase the likelihood of detecting distant homologs that are evolutionary conserved but significantly diverged.

---

Conclusion

If your initial BLAST search doesn’t identify any homologous sequences, adjusting the scoring matrix can help reveal more matches. PAM250 is typically the best choice for sequences that are expected to be divergent or come from novel sources with evolutionary differences. However, matrices like BLOSUM62 and BLOSUM80 may also be useful depending on the level of conservation in the sequence. When in doubt, experimenting with different matrices can improve the sensitivity and specificity of your sequence similarity searches.

---

Correct Answer: (c) PAM250