DIAMOND v2.2.4

• Improved the performance of the clustering workflows.
• Improved the performance of reading FASTA files.
• Added the option --single-step to perform one computational step of clustering then exit when using the --parallel-tmdir option, making it easier to orchestrate big clustering runs in a controlled way.
• Added progress output for sequence loading in clustering workflows.
• Fixed two potential issues of increased memory use in the clustering workflows.
• Added the option --self for self-alignment of the database file.
• Fixed a bug that could cause slightly reduced sensitivity on the ARM64 architecture.
• Improved performance of the search and clustering workflows on the ARM64 architecture.
• Fixed a potential Letter count not set error in the clustering workflows when run in parallel across multiple nodes.