Improved genomic island predictions with IslandPath-DIMOB
Bibliographic record
Abstract
Motivation: Genomic islands (GIs) are clusters of genes of probable horizontal origin that play a major role in bacterial and archaeal genome evolution and microbial adaptability. They are of high medical and industrial interest, due to their enrichment in virulence factors, some antimicrobial resistance genes and adaptive metabolic pathways. The development of more sensitive but precise prediction tools, using either sequence composition-based methods or comparative genomics, is needed as large-scale analyses of microbial genomes increase. Results: IslandPath-DIMOB, a leading GI prediction tool in the IslandViewer webserver, has now been significantly improved by modifying both the decision algorithm to determine sequence composition biases, and the underlying database of HMM profiles for associated mobility genes. The accuracy of IslandPath-DIMOB and other major software has been assessed using a reference GI dataset predicted by comparative genomics, plus a manually curated dataset from literature review. Compared to the previous version (v0.2.0), this IslandPath-DIMOB v1.0.0 achieves 11.7% and 5.3% increase in recall and precision, respectively. IslandPath-DIMOB has the highest Matthews correlation coefficient among individual prediction methods tested, combining one of the highest recall measures (46.9%) at high precision (87.4%). The only method with higher recall had notably lower precision (55.1%). This new IslandPath-DIMOB v1.0.0 will facilitate more accurate studies of GIs, including their key roles in microbial adaptability of medical, environmental and industrial interest. Availability and implementation: IslandPath-DIMOB v1.0.0 is freely available through the IslandViewer webserver {{http://www.pathogenomics.sfu.ca/islandviewer/}} and as standalone software {{https://github.com/brinkmanlab/islandpath/}} under the GNU-GPLv3. 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 | 0.002 | 0.006 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.005 | 0.002 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.002 | 0.002 |
| Open science | 0.001 | 0.002 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.005 | 0.003 |
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".