Fully automated sequence alignment methods are comparable to, and much faster than, traditional methods in large data sets: an example with hepatitis B virus
Bibliographic record
Abstract
Aligning sequences for phylogenetic analysis (multiple sequence alignment; MSA) is an important, but increasingly computationally expensive step with the recent surge in DNA sequence data. Much of this sequence data is publicly available, but can be extremely fragmentary (i.e., a combination of full genomes and genomic fragments), which can compound the computational issues related to MSA. Traditionally, alignments are produced with automated algorithms and then checked and/or corrected "by eye" prior to phylogenetic inference. However, this manual curation is inefficient at the data scales required of modern phylogenetics and results in alignments that are not reproducible. Recently, methods have been developed for fully automating alignments of large data sets, but it is unclear if these methods produce alignments that result in compatible phylogenies when compared to more traditional alignment approaches that combined automated and manual methods. Here we use approximately 33,000 publicly available sequences from the hepatitis B virus (HBV), a globally distributed and rapidly evolving virus, to compare different alignment approaches. Using one data set comprised exclusively of whole genomes and a second that also included sequence fragments, we compared three MSA methods: (1) a purely automated approach using traditional software, (2) an automated approach including by eye manual editing, and (3) more recent fully automated approaches. To understand how these methods affect phylogenetic results, we compared resulting tree topologies based on these different alignment methods using multiple metrics. We further determined if the monophyly of existing HBV genotypes was supported in phylogenies estimated from each alignment type and under different statistical support thresholds. Traditional and fully automated alignments produced similar HBV phylogenies. Although there was variability between branch support thresholds, allowing lower support thresholds tended to result in more differences among trees. Therefore, differences between the trees could be best explained by phylogenetic uncertainty unrelated to the MSA method used. Nevertheless, automated alignment approaches did not require human intervention and were therefore considerably less time-intensive than traditional approaches. Because of this, we conclude that fully automated algorithms for MSA are fully compatible with older methods even in extremely difficult to align data sets. Additionally, we found that most HBV diagnostic genotypes did not correspond to evolutionarily-sound groups, regardless of alignment type and support threshold. This suggests there may be errors in genotype classification in the database or that HBV genotypes may need a revision.
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 distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
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 teacher head, 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".