​The first public release of data from the HelmBank project is now live with Release R1. This release provides a new set of voucher-linked parasite DNA barcode records designed to improve how helminths are detected and identified from wildlife. This release adds 45 barcodes from 20 newly sequenced specimens with data for the markers COI, 16S, and ITS. Release R1 is part of a larger, growing, and actively curated collection of parasite specimens from wildlife. It currently focuses on the helminth parasites of Neotropical mammals, but coverage is quickly expanding to include a broader array of host taxa.

Parasites are a major—often overlooked—component of biodiversity, and they are central to wildlife health, conservation decision-making, and disease surveillance strategies (e.g., One Health). Yet accurate molecular identification depends on reference sequences that are expert-verified and relevant to local hosts and regions—resources that remain sparse for wildlife systems in the tropical Americas.

HelmBank’s goal is to translate parasite specimens collected from wildlife into high-quality reference data: DNA barcodes tied to morphological identifications, host species, and collection metadata, so that researchers and practitioners can more reliably:

• assign taxa based on sequences detected using DNA-based monitoring
• compare parasite communities across hosts and regions in molecular ecology
• support veterinary and One Health investigations at wildlife–livestock–human interfaces

​Release R1 is the first public slice of a larger curated collection. The current sample dataset includes 105 parasite specimens linked to host records spanning at least seven mammal orders—a cross-section of Neotropical wildlife that is frequently the target of conservation and wildlife health programs but yet badly under sampled for parasite diversity.

Host breadth
The working collection includes helminths from diverse mammal hosts, including:

• Wild felids: ocelot (Leopardus pardalis), jaguar (Panthera onca), cougar/puma (Puma concolor), jaguarundi (Herpailurus yagouaroundi)
• Sloths & anteaters: brown-throated three-toed sloth (Bradypus variegatus), Hoffmann’s two-toed sloth (Choloepus hoffmanni), southern tamandua (Tamandua tetradactyla), giant anteater (Myrmecophaga tridactyla)
• Large-bodied mammals: lowland tapir (Tapirus terrestris), white-lipped peccary (Tayassu pecari)
• Wild canids: crab-eating fox (Cerdocyon thous)
• Opossums: big-eared opossum (Didelphis aurita), white-eared opossum (Didelphis albiventris), gray four-eyed opossum (Philander quica)
• Armadillos: nine-banded armadillo (Dasypus novemcinctus), southern three-banded armadillo (Tolypeutes matacus), screaming hairy armadillo (Chaetophractus vellerosus), large hairy armadillo (Chaetophractus villosus)
• Livestock: domestic goat from rural Argentina (Capra aegagrus hircus)

Host breadth is intentional: it makes the resulting reference sequences more likely to improve parasite detection and identification in real-world monitoring programs—where wildlife samples come from a mix of species, landscapes, and levels of contact with people and livestock.

Parasite breadth
Across these hosts, the working collection currently spans three major helminth phyla:

• Roundworms (Nematoda)
• Flatworms (Platyhelminthes) — including cestodes and trematodes
• Thorny-headed worms (Acanthocephala)

HelmBank is being built with traceability and verification as first principles: Expert morphological identifications are captured alongside molecular data whenever possible. Ongoing curation includes taxonomic updates and corrections, including review against external resources.

Release R1 of HelmBank includes a priority set of initial "field-to-sequence" data that we can publish now, even while the broader collection continues through verification and sequencing. These data originate from 20 helminth specimens and include 45 new barcodes from COI (N = 11), 16S (N = 19), and ITS1 (N = 15).

Release R1 includes an initial subset of sequenced specimens linked to hosts that span multiple mammalian lineages of the Neotropics.

• Lowland tapir (Tapirus terrestris)
• Sloths: brown-throated three-toed sloth (Bradypus variegatus), Hoffmann’s two-toed sloth (Choloepus hoffmanni)
• Opossums: big-eared opossum (Didelphis aurita), white-eared opossum (Didelphis albiventris), gray four-eyed opossum (Philander quica)
• Armadillos: nine-banded armadillo (Dasypus novemcinctus), screaming hairy armadillo (Chaetophractus vellerosus)

Currently identified parasitic helminths associated with these hosts span two phyla: Nematoda and Platyhelminths:

• Physalopteridae (Physaloptera, Turgida)
• Kathlaniidae (Cruzia)
• Aspidoderidae (Aspidodera)
• Spirocercidae (Paraleiuris, Physocephalus)
• Trichuridae (Trichuris)
• Strongylidae (Neomurshidia)
• Spiruridae (Tejeraia)
• …and one as-yet unidentified Cestoda (Anoplocephalidae).

Several specimens in the release are resolved to species level, examples including:

• Cruzia tentaculata
• Turgida turgida
• Physaloptera papilontruncata
• Aspidodera scoleciformis

Not every record is species-level, reflecting real-world constraints on the painstaking work of identifying parasites by microscope when they vary in specimen condition, life stage, and taxonomic group. Future releases will increase resolution over time as identifications are refined and the reference library expands.

Wildlife biology & conservation: improve detection and interpretation of helminths as part of biodiversity monitoring and host health assessments.

Parasite diversity research: compare host-associated parasite communities and flag potential new lineages and host records.

One Health & veterinary applications: support parasite surveillance where wildlife, domestic animals, and humans interact.

Molecular ecology & bioinformatics: use HelmBank as a reference resource for read assignment, benchmarking, and reproducible pipelines.

If you'd like to use the data, please cite DOI dx.doi.org/10.5883/DS-HELMBR1
Guidance on peer-reviewed citation (*coming soon*)

Future releases of HelmBank will expand both host and parasite coverage beyond R1, including additional helminth groups already present in the working collection (e.g., acanthocephalans and trematodes) and new host taxa (e.g., wildcats, canids, peccaries, and anteaters) as sequences and metadata packages.

To stay updated:

• Review the HelmBank releases index for updates
• Explore the project overview
• Follow our news & updates about parasites

HelmBank exists because veterinarians, field biologists, parasitologists, and molecular ecologists are working together to connect specimens to sequence to metadata. If you are interested in collaborating—especially around under-sampled host taxa, new regions, or integration with monitoring programs—please reach out.

• Work with us
• Contact

In many dietary DNA metabarcoding studies, sampling and replication tends to be framed around predefined groups:

• Species A vs. Species B
• Dry season vs. wet season
• Treatment vs. control
• Population 1 vs. Population 2

We are taught to ask ourselves: How many samples do we need to collect per group for a statistically robust sampling design?

But what if group identity does not need to be the primary unit of analysis in the first place?

Recent analytical approaches — including the use of unsupervised and minimally supervised machine learning tools — allow ecological patterns to emerge directly from dietary data without requiring us to impose a priori sampling categories on the "groups' that we have under study. When that happens, the logic of replication changes.

Replication still matters.
But why it matters is different.

In classical statistics, we target meaningful sample sizes with replication that serves to:

• Estimate within-group variance
• Compare means or multivariate centroids
• Test differences between predefined categories

In this framework, sample size per group is central. Balance matters and power analyses often assume categorical groups are the target for sampling.
🔗 Post: How Many Samples for a Dietary DNA study?

This logic remains powerful and appropriate in many ecology and conservation contexts — especially when management decisions hinge on explicit contrasts (e.g., restored vs. degraded habitat).

• Experimental design remains one of the core skills that ecologists need to learn and practice
• Experimental manipulations remain the gold standard for identifying causal mechanisms in ecology
• Experimental field studies are among our group's favorite and most productive lines of research; we use them for understanding food webs at study sites around the world

But it is not the only analytical pathway available.

In our recent study published in Proceedings of the National Academy of Sciences, we analyzed large-herbivore diets from Yellowstone National Park using relatively simple machine learning tools.

Rather than defining groups such as “species × season” in advance, we allowed patterns in the dietary data to organize themselves. The algorithms we used identified structure based on shared dietary signals — revealing ecological gradients and overlap that outshined any categorical labels that we could have applied.
​
This does not mean species identity was irrelevant, but it does mean that structure can emerge from the data without requiring us to focus on it.

This insight really matters for study design.

When you remove predefined groupings, replication no longer needs to be a central target that we use to achieve balanced sampling of all categories.

Instead, it serves to:

• Represent the full range of ecological variation present in a system
• Capture dietary heterogeneity across individuals, regardless of a priori grouping(s)
• Cover multidimensional dietary space
• Stabilize pattern detection in high-dimensional data

In this context, the central question shifts from:

“How many samples do I need to find per group?”

to:

“How well does my sampling capture the ecological variation that exists out there in nature?”

Replication is still essential — but its purpose is representational rather than categorical.

There is a subtle but important distinction in these concepts.

In group-based sampling designs, you pay attention to:

• Equal replication across treatments
• Sufficient power to detect differences

In any structure-detection framework, you pay more attention to:

• Coverage of environmental gradients
• Inclusion of rare but informative dietary signals
• Adequate sampling across behavioral diversity

An imbalanced dataset that you might worry about in a classic experiment can still become extremely informative if it adequately spans ecological space. Conversely, perfectly balanced groups can still miss important gradients if sampling is narrow.
​
The emphasis shifts from symmetry to coverage.

The distinction between representation and balance does not mean:

• Sample size no longer matters
• Small datasets are good enough
• Machine learning eliminates the need for thoughtful sampling design
• The field is moving beyond the need for mechanistic experiments to understand nature

High-dimensional dietary data can be noisy, and sparse sampling can produce unstable clustering or misleading gradients. Overinterpretation remains a risk, especially if find ourselves using data-hungry approaches without enough raw data to feed them.

Unsupervised approaches require at least as much planning and careful consideration as traditional comparisons — the key distinction is all about what we are replicating and how it supports inference.

Structure-first approaches can be particularly useful when:

• Group boundaries are biologically fuzzy
• Species groups overlap extensively in diet
• Environmental gradients are continuous rather than discrete
• You suspect hidden structure not captured by simple categories

In Yellowstone National Park, for example, dietary overlap among large herbivores is shaped by shared landscapes, seasonal dynamics, and plant community structure.

​Allowing patterns to emerge from the data helped reveal how ecological organization did not always align predefined groupings.

Even in structure-detection studies, replication helps lead us toward reliable inferences.

More samples:

• Better represent ecological space
• Improve detection of subtle gradients
• Reduce sensitivity to outliers
• Strengthen generalizability

In many practical cases, replication levels similar to those recommended for comparative designs should be the target (e.g., ~20–30 independent samples per ecological unit). This is not because groups must be balanced, but because ecological variation must be adequately sampled.

The logic evolves, but the need for replication does not disappear.

The broader takeaway is not that predefined groups are obsolete.

It is that dietary DNA metabarcoding now supports multiple analytical frameworks:

• Hypothesis-driven comparisons
• Gradient-based inference
• Clustering and dimensionality reduction
• Hybrid approaches combining categorical and emergent structure

Thoughtful study design requires aligning replication with the type of inference you intend to make.

If your management question hinges on specific contrasts, group-based replication remains essential.

If your goal is to detect latent structure or ecological gradients, prioritize coverage of variation across individuals and environments.

In both cases, replication fuels our inference — it doesn't lead us to the truth about nature on its own.

As dietary DNA datasets grow and analytical tools diversify, we may see a shift from strictly categorical thinking toward more flexible representations of ecological structure.

That shift does not reduce the importance of sampling.

• It raises the bar for intentional study design.
• Replication is no longer only about statistical power between bins.
• It is about faithfully representing ecological reality.

And that remains the central goal in our efforts to achieve impact in conservation through molecular ecology.

Designing a dietary DNA metabarcoding study often begins with a deceptively simple question: How many samples do I really need to collect?

There is not a universally “correct” number. We all want to have a large enough sample size for a powerful analysis. But it can be extremely challenging to collect fresh scat samples from wild animals—especially when they are rare and widespread—and then we face the cost of analyzing what we get.

To answer this question, we need to focus mostly on the ecological inferences we want to make. Are we trying to compare groups? Estimate niche breadth? Detect rare food items? Describe seasonal shifts? The number of samples required to detect differences between sample sets is often very different from the number needed to perfectly catalog everything in a diet. So, I want to share some helpful rules of thumb based on experience across a wide variety of study systems...

Several factors shape how many samples you should target for collection:

• Individual-level diet variation
• Population-level diet heterogeneity
• Temporal variability
• Diet diversity (specialist vs. generalist feeders)
• Whether your goal is description or comparison

Highly generalized feeders with diverse prey taxa typically require more replication than specialists with narrow and relatively constant diets. Systems with strong seasonal shifts may require replication across time. Populations occupying heterogeneous landscapes may show greater variance among individuals that can only be quantified with effort.

The most important question is not “How many samples is enough?” but "When does the ecological signal that I need to detect become robust?"

It is much easier to detect differences between diets than to perfectly catalog everything in a diet.

If your goal is exhaustive description—identifying every taxon consumed and estimating its relative contribution—replication requirements can be high, especially in species with diverse diets. Instrumental error becomes a significant concern as well.

If your goal is comparative or experimental—such as:

• Restored vs. degraded habitat
• Dry vs. wet season
• Species A vs. Species B
• Treatment vs. control

—then structured differences often emerge with fewer samples.

Comparative research designs are powerful because they focus on relative differences rather than perfection. In many conservation contexts, that distinction is critical. Management decisions often hinge on contrasts, and prior information is almost always limited.

As a general guideline, when I'm asked to recommend a target sample size I usually suggest aiming for 20–30 independent samples per “group.”

A “group” should be defined relative to the goals of the study, such as:

• A species
• A species × season combination
• A population
• A treatment
• A site

One common strategy that people hear about to reduce costs involves combining samples from multiple individuals into a pooled composite sample. I generally recommend against this approach for dietary DNA studies.

Pooling masks inter-individual variation, and that variation is statistically powerful—even when total sample sizes are modest. Quantifying differences among individuals is often essential for revealing ecological structure, niche partitioning, or behavioral flexibility that would otherwise remain hidden.

Even relatively small numbers of independent samples frequently provide more inferential leverage than a few pooled composites.
​
In dietary DNA research, independent replication is far more valuable than artificial composites.

My take: You should analyze them. Does that appear to be a little cavalier? Perhaps. But do it anyway. Here’s why…

A small dataset can still be valuable, particularly when:

• Your system is understudied
• Your work is at an exploratory or hypothesis-generating stage of development
• Your samples were so difficult to obtain that no one is likely to try again anytime soon
• You are in a unique position to publish diet profiles that will be new to the literature

It is worth noting that not all dietary DNA studies need to rely on predefined groups such as species or treatment categories. In some cases, analytical approaches—particularly unsupervised or minimally supervised machine learning—allow ecological structure to emerge directly from the data without imposing a priori bins. In our recent work in Yellowstone, for example, we used data-driven methods to identify structure in large herbivore diets without defining groups in advance. In analyses like these, the question shifts from “How many samples per group?” to “How well does our sampling capture the underlying ecological variation?” Replication still matters, but it is used to ensure representation of ecological variation rather than to balance sampling across predefined categories. This distinction is subtle but important: thoughtful sampling strategies remain essential, even when group identity is not the central organizing principle of the analysis, but pre-defining target sample sizes may not always be required.

🔗 Post: Do You Even Need “Groups”? Rethinking Replication in Dietary DNA Studies.
🔗 Software & Data: Our DNA metabarcoding pipelines and code are freely available for use.

If you can, it may be wise to consider a pilot study before scaling up. You can:

• Generate rarefaction curves
• Estimate among-individual variance
• Evaluate precision in the data
• Assess detection consistency

Pilot data can help you develop your final target sample size so you don't over- or under-sample.

It used to be hard to do this due to the cost and labor involved in putting together a full Illumina run: by the time you had enough samples to justify the cost of the run, you might as well complete the whole project in one go. (Sometimes if you knew the director of a core facility, they might be willing to ‘spike’ a few of your samples into a run that somebody else was paying for… but that wasn't always an option.)

Portable sequencers now make it much more cost-effective to run small pilots quickly. Our group, through the Genomic Opportunities Lab, may be able to help you create pilot dietary DNA data if you are an academic and conservation practitioner.

There is no single “correct” number of samples for a dietary DNA metabarcoding study.

Thoughtful replication—matched to the scale of inference—matters much more than achieving your maximal replication. In conservation research especially, study design should prioritize the ability to detect meaningful ecological differences over the pursuit of exhaustive completeness. We need to get that part of the study right or we risk eroding trust in our credibility. So design your sampling strategy around the question you are trying to answer. The rest follows.

Jump to: Rules | Risks | Reasons for Concern | University Links & Policies

Artificial intelligence is increasingly useful as a tool to improve our research and learning. We use it to troubleshoot code, polish writing, get good ideas about how to visualize data, create document templates that save time on busywork… But at the same time, we must be cognizant of legitimate concerns about the accuracy of information it can provide, its ability to reuse confidential information that we disclosed in chats, and the risk of short-circuiting our own creative uses of the scientific method.

This post summarizes rules that lab members should follow when using AI in their work. I do not want to regurgitate the types of dry, legalese we are provided by our employer—rather I will attempt to illustrate the fine-line we have to walk to ensure we are using the tool appropriately while minimizing the risk of unintended harm. I will summarize reasons for concern using language familiar to biologists and conservationists broadly. Some of the details are specific to researchers at Brown, but I believe the information is readily transferable and I welcome others to use this document as a template for their own policies.

​Please read on... 

1.     Never upload or paste your data, code, or scientific writing into a commercial AI platform. Doing so would risk violating legal requirements concerning confidentiality and the use of original data, while compromising our collective efforts to advance the bounds of knowledge.
 
2.     Always use Brown University’s internal LibreChat system if you want to use AI to check your work or save you time. This system provides access to many of the same AI tools that researchers often access online—ChatGPT, Claude, Canva AI, etc.—but it does not share our data with external tech companies.
 
3.     If you use commercial AI platforms—including Google Search with AI—do so with extreme care to ensure you do not violate Rule #1. Only do so if you can be certain that you are entering completely generic prompts that cannot be connected to your work in the lab—if there is any doubt, use LibreChat instead.
 
4.     Be able to independently verify all insights and information you obtain from these platforms when asked. The baseline quality of information these platforms can provide is limited based on the quality of information sources they are using—they make mistakes and have the potential to mislead with some frequency. It is worth remembering that our job is to generate information that no one has ever had before and thus you cannot rely on these tools for information that you are not independently verifying yourself.

​Read each of the four pairs examples and consider differences between each query is constructed:

​In each of the first cases, you see that there are ways to create prompts that enable you to use these kinds of algorithms as time-saving tools without compromising the confidentiality of your work in the lab. 

In each of the second cases, you see that it is all too easy to provide third-party commercial platforms with information that you may not be allowed to knowingly disclose—even if that is not your intent.

This is the difference between using AI as a tool to accelerate our research and get past roadblocks versus using AI to do things for us. The second case clearly has the potential to cause frustrating experiences and lead to errors—reason enough not to outsource your responsibility to a commercial chatbot—but that is just about the user's short-term experience. The risk of making unintended disclosures can be far more significant and long-lasting.

Below, I will illustrate reasons for concern. Awareness will help ensure we are able to use the tools appropriately and are quick to recognized inappropriate types of interactions.

Disclosing confidential, protected information. We work with protected plants and animals; ethical considerations, and sometimes legal requirements, prevent us from disclosing sensitive information that could jeopardize them, their habitats, or the people who work to protect them. For example, we redact or coarsen information about where we sample protected wildlife populations when we publish results in order to avoid inadvertently aiding poachers. However, AI platforms store information you provide in order to provide information to others—this is a significant concern. Brown also has categories of data risk that may apply: See the "Data Risk Classifications" link below.
 
Getting scooped, plagiarism, missing out on credit. We have ethical obligations to one another, and to our funders, to ensure the utmost integrity in our research. We strive to share all scientific information quickly and responsibly—and especially with respect to tax-payer funded work we are required to do so—but we undermine our own best efforts when our work is shared with others before we can fully verify results with our careful quality control procedures. This can happen, even unknowingly, if another user enters a prompt and the response they get is based on unpublished information originating within our group. Other individuals may use that information as if it were original or their own, with no opportunity to verify it using the peer-reviewed literature or to properly credit us for the idea. In some cases, AI bots can search and find data or code and package it in ways that make it all too easy for someone else to claim credit. See the related link below about "IP" (Intellectual Property) at Brown.

Failing to verify results, perpetuating errors. Each of us bears significant responsibility for ensuring the accuracy of our results—inclusive of how our data are recorded, how samples are handled, and how data are analyzed, visualized, interpreted and communicated. Failure to do so can result in irreparable harm to our reputation, careers, and field writ large. Mistakes happen in research, but we can only correct them if we are attentive to detail—and the practice of attending to detail requires time. Failing to verify code, graphics, citations, or translations of original text are just a few of the ways that overreliance on these tools introduces the risk of perpetuating damaging errors.
 
Social and environmental harms. The explosion of AI data servers is taxing local electric systems, water supplies, and the social fabric of surrounding communities all over the world. Many of these facilities are being constructed in regions with few environmental protections, exacerbating the associated environmental harms. Life is not without impact—we should use all the tools available to maximize the quality of our science and progress in our careers—but we should be cognizant of impacts and strive for moderation.
 
Failure to learn, metacognition. In science and academia, your brain is your greatest asset. It is what provides you with genuine intelligence. Evidence shows that we learn best by recalling information and applying it in new situations. What may feel like a time-saving opportunity that advances your work could ultimately undermine your goals if it becomes a tactic to avoid investing sufficient effort in the difficult task of learning and improving with time. Therefore, I recommend learning about the “Desirable Difficulty” concept—it will make you a better student and teacher, who is prepared to use AI effectively while managing its associated risks. It is ultimately our job to recognize what advice we get from AI is based on useful information, and learning to do this efficiently and effectively is going to become an increasingly important part of academic training. 

Additional links relevant to the use of AI at Brown University:

• Intellectual Property: https://policy.brown.edu/policy/copyright-ownership-and-use-policy 
• Data Risk Classifications: https://it.brown.edu/policies/data-risk-classifications
• Sheridan Center Teaching with AI in Mind: https://sheridan.brown.edu/resources/teaching-ai-mind
• University AI Usage Guidance: https://provost.brown.edu/communications/potential-impact-ai-our-academic-mission
• Documentation and Privacy Statement: https://docs.ccv.brown.edu/ai-tools
• Brown Technology Innovations: https://bti.brown.edu/

Working with dietary DNA metabarcoding data? Unsure how to concisely summarize your workflow for publication? Tired of all the effort required to format your data tables for archiving in Dryad, supplementary materials, or other archives? The lab has posted new code to our GitHub repository that will help you solve all of these problems.

We have posted detailed new protocols describing our methods to sequence key mammalian DNA barcodes. They can be found together with a growing number of field and lab protocols on the Kartzinel Lab's centralized protocol page.

You will find protocols for both the D-loop of the mitochondrial control region and the 16S marker are useful for identifying a diversity of mammals, and can be routinely amplified from degraded material such as fecal DNA. We have frequently used these protocols to confirm the identity of mammals in studies involving dietary DNA metabarcoding and/or host-microbiome interactions. They are also very useful for phylogenetic analyses. We have used various polymerases over the years, so these protocols may depart slightly from previously published versions (e.g., Kartzinel et al. 2019 PNAS). However, they reflect our current state-of-the-art strategy for routine work and should be generally more cost or time effective as a result of the changes. 

New feature on our Software & Data repository page: Hot off the press! Featuring code from Hannah Hoff's 2025 PNAS paper, The Apportionment of Dietary Diversity in Wildlife.

This paper presented a potentially paradigm-shifting strategy to quantify and characterize the number of unique 'diet types' that exist within a population or community. The strategy is based on a simple machine-learning algorithm and described in the Hoff et al. 2025 PNAS paper, which used the community of migratory large mammalian herbivores -- such as bison and elk -- as a prime example.

The strategy makes use of dietary DNA metabarcoding data and extensive local plant DNA reference library to characterize variation in animal diets. It then applies a machine-learning algorithm called "Partitioning Around Medoids" -- or "PAM" -- to help organize samples into an optimal number of clusters to maximize variation between groupings. This allows us to recognize patterns in the data, without having to apply a priori assumptions about the groupings (e.g., should we lump all samples from a species together?). 

After clusters are identified, the code also presents a strategy for performing Indicator Species Analysis to identify plant taxa that contribute strongly to the clustering pattern we observed.

You can link to the open-source code and data repositories are at:

• Dryad https://doi.org/10.5061/dryad.mgqnk99b8
• Zenodo https://doi.org/10.5281/zenodo.15634157

Featuring our pipeline as a set of open-source repositories, hosted by GitHub and organized for easy access on our webpage. The pipeline is optimized for use on Oscar, which is the Brown University cluster, but it can be adapted readily to other institutions. We also welcome collaborators, who can formally gain access to our ready-to-run infrastructure on Oscar when they train or collaborate with us -- something we have been successfully making easier and more affordable for external collaborators to do!

An overview of our bioinformatic pipeline is provided below, current as of January 2026, and maintained on our GitHub site. 

Since June 2025, we have increasingly made our internal lab methods publicly visible on the "Protocols" section of our webpage. We began with some of the most frequently requested protocols that speak to the unique strengths of our lab's work and experience, featuring field-to-lab protocols for collecting and banking dietary samples, parasite samples, and plant barcode samples. We have expanded to include...

several protocols for plant barcoding and metabarcoding, which are among the most widely used in our lab as well as those that we collaborate with. 

More protocols will be posted continuously, including sequencing protocols on various platforms (Sanger, Nanopore, Illumina).

Protocols are being posted in both English and Spanish, especially in cases where protocols are being used in primarily Spanish-speaking countries. 

Please comment or contact the PI if there is a protocol you would especially like to see, or if you have any suggestions for improvements or alternatives. A number of our protocols used to be publicly available via our lab wiki, but unfortunately the site that hosted us is no longer active and so we are transitioning. 

Featured Software from the Kartzinel Lab: Geographic Coverage of DNA Barcodes. The inaugural code repository to be highlighted in our Featured Software section of the Software & Data page presents the Quarto Code Book published in association with our Molecular Ecology Review Paper, "Global Availability of Plant DNA Barcodes as Genomic Resources to Support Basic and Policy-Relevant Biodiversity Research" can be easily modified to evaluate the geographic coverage of other data sets. Although the featured code emphasizes geographic coverage from our work in Yellowstone National Park...

The codebook itself is readily customizable for a wide variety of uses. To facilitate re-use, we have divided it into four parts that follow the Methods section of the publication and includes eight code notebooks that we used in the analyses for this publication. Source code is available in a GitHub repository

To see how data are formatted and work through the Yellowstone example as a vignette, supplementary data sets (S1-S4) can be downloaded from the supplement available with the published paper. Move those four files into the /data folder in the GitHub repository when running the code notebooks. If running notebooks from the repository, headers in the data retrieved may differ than headers in the datasets provided in the Supplemental Materials.

The sections documenting the workflow are as follows: 
Building global BOLD data

1. Building the BOLD dataset

Geographic coverage

1. Coverage by climatic zones
2. Coverage by country

Taxonomic coverage

1. Fetching data from ITIS
2. Correlations – plant species and available barcodes

Case-study Yellowstone National Park

1. Continent-scale barcode coverage of lodgepole pine and big sagebrush
2. Fetching data from GBIF
3. Case-study in geographic coverage of site-based reference data