SPANNER: taxonomic assignment of sequences using pyramid matching of similarity profiles
Bibliographic record
Abstract
BACKGROUND: Homology-based taxonomic assignment is impeded by differences between the unassigned read and reference database, forcing a rank-specific classification to the closest (and possibly incorrect) reference lineage. This assignment may be correct only to a general rank (e.g. order) and incorrect below that rank (e.g. family and genus). Algorithms like LCA avoid this by varying the predicted taxonomic rank based on matches to a set of taxonomic references. LCA and related approaches can be conservative, especially if best matches are taxonomically widespread because of events such as lateral gene transfer (LGT). RESULTS: Our extension to LCA called SPANNER (similarity profile annotater) uses the set of best homology matches (the LCA Profile) for a given sequence and compares this profile with a set of profiles inferred from taxonomic reference organisms. SPANNER provides an assignment that is less sensitive to LGT and other confounding phenomena. In a series of trials on real and artificial datasets, SPANNER outperformed LCA-style algorithms in terms of taxonomic precision and outperformed best BLAST at certain levels of taxonomic novelty in the dataset. We identify examples where LCA made an overly conservative prediction, but SPANNER produced a more precise and correct prediction. CONCLUSIONS: By using profiles of homology matches to represent patterns of genomic similarity that arise because of vertical and lateral inheritance, SPANNER offers an effective compromise between taxonomic assignment based on best BLAST scores, and the conservative approach of LCA and similar approaches. AVAILABILITY: C++ source code and binaries are freely available at http://kiwi.cs.dal.ca/Software/SPANNER. CONTACT: beiko@cs.dal.ca SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.003 | 0.012 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.005 | 0.004 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.002 | 0.004 |
| Open science | 0.002 | 0.003 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.016 | 0.010 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".