Lichen Biodiversity Informatics
Nov 2024 – Present · Arizona State University Research Affiliate
Paused — resuming end of summer 2026
Overview
Collaboration with Dr. Frank Bungartz, Collections Manager of Lichens at Arizona State University and domain authority on lichen taxonomy. A multi-component research program spanning database engineering, phylogenetics, and deep learning image classification for the Consortium of Lichen Herbaria (CNALH).
ML Image Classification Pipeline
BioCLIP ViT-L/14 + ArcFace metric learning classifier for species-level lichen identification from photographs.
- Two-stage training: large-scale noisy pretraining on iNaturalist aggregated imagery → fine-tuning on expert-labeled Consortium data (~15,000 images, ~10,000 species)
- UMAP + HDBSCAN clustering over LIAS DELTA morphological data (10,709 species, 880 chemical compounds)
- Clustering revealed clean photobiont-type separation — validating signal in morphological feature space
- Embedding ensemble: BioCLIP / DINOv2 / CLIP
Phylogenetic Analysis Pipeline
End-to-end pipeline from raw GenBank HTML-wrapped flat files through maximum likelihood phylogenetic inference:
- Parsed ~51K GenBank DNA records (8 loci: ITS, 18S, 28S, RPB2, and others) — 49,610 valid records
- MAFFT alignment, trimAl trimming, IQ-TREE2 maximum likelihood inference
- Primary focus on Lecanoromycetes phylogeny (ITS, 18S, 28S loci)
- Mapping LIAS morphological and chemical trait data onto inferred trees
Database Engineering & Taxonomy Cleanup
PostgreSQL database (fungix schema) integrating LIAS DELTA, Mycobank, and Index Fungorum:
- Removed 38,753 non-lichen taxa from the Consortium database
- Added 68,717 new validated taxa from Mycobank and Index Fungorum
- Updated 162,465 taxa records with external source references and protolog data
- Contributor to Symbiota open-source platform — sourceIdentifiers across all taxonomic tree explorers
Presentations
Taxonomic Data Enhancement of the Consortium of Lichen Herbaria (Symbiota)
ABLS Consortium Meeting · April 2026