Supplemental Material for Gabidulina and Rudman, 2025

Collecting phenotypic data from many individuals is critical to answering fundamental biological questions, particularly in genetics. Yet, whole organismal phenotypic data are still often collected manually; limiting the scale of data generation, precluding reproducible workflows, and creating potential bias. Computer vision can ameliorate these issues, but currently available packages only operate with specific inputs and have limited scalability. We present Machine Learning Data Acquisition for Assessing Population Phenotypes (MLDAAPP), a package of tools built on YOLOv8 for collecting phenotypic data from groups of individuals that is flexible for generating counts (e.g. <em>Drosophila</em> fecundity, census), movement (e.g., activity, aversive behavior), and many other potential phenotypes. MLDAAPP is both accurate and uniquely effective for obtaining phenotypic data in challenging conditions - particularly images and videos of varying quality derived from both lab and field environments. Employing MLDAAPP solves key issues of reproducibility and increases the scale and scope of data generation.  
<br>
This dataset contains: Videos &amp; Photos that were tested in the main body text; Both AI annotated and Original sources. For more details see maintext or head to  ganamir/MLDAAPP.
<br>