Interpreting the <scp>flock</scp> algorithm: a reply to Anderson & Barry (2015)
Bibliographic record
Abstract
Anderson & Barry (Molecular Ecology Resources, 2015, 10, 1020-1030) compared a reprogrammed version of flock (Duchesne & Turgeon , Molecular Ecology Resources, 2009, 9, 1333-1344), flockture, to a particular model of structure (Pritchard , Genetics, 2000, 155, 945-959) that they propose is equivalent to flock, a non-MCMC, non-Bayesian algorithm. They conclude that structure performs better than flockture at clustering individuals from simulated populations with very low level of differentiation (FST c. 0.008) based on 15 microsatellites or 96 SNPs. We rather consider that both algorithms failed, with proportions of correct allocations lower than 50%. The authors also noted the slightly better performance of flockture with SNPs at intermediate FST values (c. 0.02-0.04) but did not comment. Finally, we disagree with the way the processing time of each program was compared. When compared on the basis of a run leading to a clustering solution, the main output of any clustering algorithm, flock, is, as users can readily experience, much faster. In all, we feel that flock performs at least as well as structure as a clustering algorithm. Moreover, flock has two major assets: high speed and clear, well validated, rules to estimate K, the number of populations. It thus provides a valuable addition to the set of tools at the disposal of the many researchers dealing with real empirical data sets.
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 | 0.027 | 0.125 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.002 | 0.002 |
| Bibliometrics | 0.002 | 0.002 |
| Science and technology studies | 0.006 | 0.017 |
| Scholarly communication | 0.008 | 0.013 |
| Open science | 0.008 | 0.004 |
| Research integrity | 0.039 | 0.063 |
| Insufficient payload (model declined to judge) | 0.007 | 0.012 |
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".