szabgab/ullar

An ULtrafast, scaLable, Accessible, and Reproducible phylogenomic pipeline.

ullar

ULLAR is an Ultrafast, scaLable, Accessible, and Reproducible pipeline for phylogenomics. Why do need another pipeline for phylogenomics? Our main reasons:

Performance and Scalability

1. Most pipelines are written in scripting languages (e.g., Python, R) and are not optimized for performance. While for most pipelines, the heavy lifting is done by external tools (often are fast) the overhead of the pipeline itself can be significant, especially when some work is done using the pipeline internal code. ULLAR is aimed to minimize this overhead.

2. Most pipelines are not optimized for parallelism. Whenever possible we want to take advantage of multi-core systems. We will do it safely without data race.

Accessibility and Reproducibility

1. ULLAR's goal is to be easy to install and easy to use. We want to minimize dependencies, avoid containerization (e.g., using Docker or Singularity), and make the pipeline as self-contained as possible. Whenever possible we will interact with available tools using Foreign-Function Interface (FFI), so the tool will be contained within the ULLAR binary. We will also simplify command arguments and will accompany the pipeline with comprehensive documentation. This way, non-tech-savvy users or those without technical support can easily install the pipeline and efficiently conduct phylogenomic analyses.

2. ULLAR will support Windows for part of the pipeline. Most of day-to-day user are using Windows. We understand that many bioinformatic tools are not available for Windows. Nonetheless, we wanted to make sure that ULLAR code is cross-platform. We will also prioritize tools that are available in Windows. We believe it will improve reproducibility and accessibility.

3. ULLAR will not only be reproducible through log output and runner scripts, but will also provide check point to resume the pipeline. No more rerunning the whole pipeline when it fails in the middle or do manual hacks to resume the pipeline.