How we got here
When this project began, we anticipated quickly setting up a few standard methods before moving on to data collection. However, we soon discovered that tools we initially considered straightforward —such as DNA extraction kits and library preparation methods—had a far greater impact than expected. Early in the process, we encountered problems  with DNA isolation kits that were not optimized for our sample types. It became evident that, to achieve meaningful results, we’d need to carefully examine each step and select the most suitable tools for each part of the workflow.

Why dogs?
Why did we choose to study dogs, specifically the Hungarian Pumi breed? Dogs are not only beloved companions, but their gut microbiome has remarkable similarities to ours. The microbiome— the diverse community of bacteria, fungi, and other microorganisms living in the gut—plays a vital role in health, influencing digestion, immune function, and even mental well-being. Understanding it better could have a major impact on health research for both animals and humans.

This project serves as an introduction to a series of ongoing studies involving nearly 100 Pumi dogs as well as human patients. By focusing on a genetically uniform group of dogs, we are able to create a model that provides insights applicable to human health. Some of these dogs were already included in this study as controls to give us more reliable baseline data. However, in this study, our primary subject was Toto,  my own Pumi (now 16.5 years old), from whom  I could readily collect samples throughout the project.

Lab setbacks and personal discoveries
Our biggest surprise was realizing how differently each DNA extraction kit performed. For example, the Zymo Research kit was excellent for DNA quality and fragment length in dog stool samples, while the Invitrogen and Macherey-Nagel (MN) kits offered their own unique strengths. These differences reinforced that there isn’t a one-size-fits-all approach in microbiome research. Each step in the process had to be considered carefully, even if it meant going back to the drawing board a few times.

This is also where we started developing minitax, a bioinformatics tool that helped standardize the data analysis across different sequencing platforms. This wasn’t just a nice-to-have tool; it became essential. Minitax allowed us to keep our analysis consistent, even as we worked with both short-read and long-read sequencing data. Having a tool that could bridge these differences made a big difference in our workflow.

Our key findings and practical advice
After running all these tests, here is what we would recommend based on different project needs:

• Affordable option: Zymo DNA isolation with Oxford Nanopore Technologies'  V1-V9 library prep is a good choice for studies that need to keep costs down.
• Balanced approach: Invitrogen DNA isolation with Illumina V3-V4 libraries offered a balance between cost and accuracy.
• High-precision profiling: For the highest detail, we found that MN DNA isolation with Illumina  whole-genome sequencing yielded the most detailed microbial profiles.

When it comes to bioinformatics, minitax worked across different sequencing types, while tools like Emu and Sourmash were better suited to specific needs, like amplicon or metagenomic sequencing.

Why this matters
So, what does all this mean? Microbiome research relies on accuracy, especially when considering applications in health and disease. However, without standardized methods, results can vary significantly, making comparisons across studies challenging. When drawing conclusions, researchers must consider that differences may stem from variations in lab procedures and bioinformatics approaches as much as from the biological data itself. We hope our work can serve as a resource to help researchers choose the right tools for their specific samples and research questions.

Looking ahead
This study highlighted the importance of standardization in microbiome research. Our findings underscore that no single solution is ideal for every sample type. Instead, we need adaptable workflows and robust tools  to handle differences in sequencing data. We hope this work will support researchers facing similar challenges, making microbiome profiling more reliable.