That is a large amount of sequences for both IQ-TREE and Nextstrain’s Auspice visualization. See this post for discussion on why it is failing, and this post for a reasonable sample size (generally 4,000 ~ 10,000). To obtain a smaller sample size, you can still start with all 78,000 samples and adjust subsampling options to preserve diversity and coverage across locations over time. This can be done with subsampling and builds.<build_id>.subsampling_scheme in your configuration .yaml file – see the genomic configuration surveillance tutorial for an example.
1 Like