Error in reduceSimMatrix for pattern 4: Scores vector does not contain all terms in the similarity matrix

[sarahtanja]

Summary

When running the semantic similarity reduction with rrvgo's reduceSimMatrix() for the interaction gene pattern 4 (at lines ~1143-1149 in code/08_go/08_GO.qmd), an error occurs:

Error in reduceSimMatrix(simMatrix.4, scores = scores.4, threshold = 0.8, :
  Scores vector does not contain all terms in the similarity matrix

Details

• The analysis works as expected for patterns 1-3, but fails on pattern 4 (which only contains a single gene/GO term) when calling reduceSimMatrix().
• The error message suggests a mismatch between the scores vector and the similarity matrix terms.
• Upstream, the GO enrichment for pattern 4 is performed on only one gene, which may be the root cause.

Steps to Reproduce

1. Run the code block at lines 1143-1149 in code/08_go/08_GO.qmd.
2. Error is thrown when calling reduceSimMatrix(simMatrix.4, scores = scores.4, threshold = 0.8, orgdb = "org.Ce.eg.db").

Expected Behavior

• The function should either handle the edge case of a single GO term gracefully (e.g., return the term itself), or provide a more informative message.

Environment

• File: code/08_go/08_GO.qmd
• R packages: rrvgo, goseq, dplyr, etc.
• Input: Only 1 gene/GO term in pattern 4.

Possible Solution

• Add an explicit check for the number of terms in the similarity matrix or scores before calling reduceSimMatrix(), or handle single-term cases upstream.

Screenshots/Error Output

Error in reduceSimMatrix(simMatrix.4, scores = scores.4, threshold = 0.8, :
  Scores vector does not contain all terms in the similarity matrix

---

This issue tracks the failure of GO term reduction for a single-gene DGE pattern. See lines 1143-1149 in the analysis notebook for details.