Use mmseqs2 MSA Generation

The amino acid sequences corresponding to the proteins you want to generate an MSA for. This will not generate a seperate MSA for each provided protein but will instead generate a single paired MSA using all provided sequences. NOTE: This service uses the ColabFold mmseqs2 API and will share your input sequences with that API. If you are not confortable or authorized sharing your sequences with the third party API, do not use this service.

Choose a mode for MSA generation: 'unpaired' creates individual MSAs for each sequence; 'paired' generates MSAs for related sequences to preserve inter-sequence relationships; 'unpaired_paired' combines both approaches, maximizing coverage by first pairing sequences, then supplementing with unpaired MSAs if needed. For more information see this blog post: https://neurosnap.ai/blog/post/6711802c8bedccdda9e14fe4

This parameter determines the minimum percentage of sequence coverage required for a sequence to be included in the multiple sequence alignment (MSA). Specifically, it refers to the proportion of residues in the query sequence that must align with residues in a potential sequence match. For example, with a coverage threshold of 50% (the default), at least half of the positions in the query sequence must have aligned positions in the matched sequence. This ensures that only sequences with sufficient overlap in the alignment space are included, which can be especially useful in filtering out shorter or partial matches that may not provide meaningful conservation information. Adjusting this parameter can fine-tune the sensitivity of the MSA, balancing the trade-off between the amount of information retained and the level of similarity across sequences.

This parameter sets the minimum percentage of sequence identity required for a sequence to be included in the MSA. Sequence identity refers to the exact match of amino acid or nucleotide residues at aligned positions between sequences. With an identity threshold of 90% (the default), only sequences with at least 90% identity to the query sequence are retained. This setting is critical for filtering out highly divergent sequences that may not accurately represent conserved evolutionary information or functional relevance. Lowering the identity threshold increases diversity within the MSA, potentially capturing more distant homologs, whereas a higher threshold ensures that only close homologs, which are more functionally similar, are considered.

Maximum number of sequences to include in the final MSA. If the generated MSA exceeds this number, it will be trimmed down to this limit (default: 2048).