Research

Lichen Biodiversity Informatics & ML Classification

Nov 2024 – Present · Arizona State University Research Affiliate

Collaborating with Dr. Frank Bungartz, Collections Manager of Lichens at Arizona State University and domain authority on lichen taxonomy, on a multi-component research program spanning database engineering, phylogenetics, and deep learning image classification.

ML Image Classification Pipeline

Developing a BioCLIP ViT-L/14 + ArcFace metric learning classifier for species-level lichen identification from images. Two-stage training design: large-scale noisy pretraining on iNaturalist aggregated imagery, followed by fine-tuning on expert-labeled Lichen Consortium data (~15,000 images across ~10,000 species).

UMAP + HDBSCAN clustering pipeline built over LIAS DELTA morphological data (10,709 species, 880 chemical compounds)
Interactive three-panel browser-based cluster viewer for exploratory analysis
Clustering analysis revealed clean photobiont-type separation — validating signal in morphological feature space
Embedding ensemble strategy: BioCLIP / DINOv2 / CLIP for robust representation

Phylogenetic Analysis Pipeline

End-to-end pipeline from raw GenBank HTML-wrapped flat files through aligned multi-locus phylogenetic inference:

Parsed ~51K GenBank DNA records (8 loci: ITS, 18S, 28S, RPB2, and others) producing 49,610 valid records
MAFFT alignment, trimAl trimming, IQ-TREE2 maximum likelihood inference
Primary focus on ITS, 18S, and 28S loci for Lecanoromycetes phylogeny
Mapping LIAS morphological and chemical trait data onto inferred phylogenetic trees

Database Engineering & Taxonomy Cleanup

PostgreSQL database (fungix schema) integrating LIAS DELTA, Mycobank, and Index Fungorum data sources:

Removed 35,000 non-lichen taxa from the Consortium database
Added 60,000 new validated taxa from Mycobank and Index Fungorum
Updated all 200,000 taxa records with external source references and protolog data
Contributor to Symbiota open-source platform — sourceIdentifiers across all taxonomic tree explorers

Parallel Investigation: PCG Classification

Active investigation of heart sound (phonocardiogram) classification using mel-spectrogram + Vision Transformer architectures and pretrained audio foundation models (HuBERT, Wav2Vec 2.0) — directly extending ECG signal processing expertise to acoustic cardiac signals.

Affiliations