Enhancing the diversity and inclusion of the life sciences workforce has become an important problem as highlighted by many organizations in the US, including NIH, NHGRI, and NSF. People with visual impairments are one of the groups that face barriers to access to the biology workforce. To overcome this challenge, it is important to understand their current barriers in biological research and education. The most common assistive technology used by people with visual impairments is the screen reader (45.2%). However, multiple studies found that many websites largely fail to meet accessibility guidelines, making it challenging or even impossible for screen reader users to access existing resources. To help gain better insights into how well people with visual impairments can access existing biological resources, we evaluated the digital accessibility of two essential resources for data-driven studies—data portals and journal websites. Using an automated evaluation tool, we collected accessibility evaluation data for a large corpus of resources (N=3,943). In addition, we collected metadata of individual resources (e.g., geospatial, temporal, and impact score data) for a more insightful analysis. All datasets, as well as the entire source code, are available online on Zenodo and GitHub under a CC-BY and MIT license, respectively.