Dear SPAdes developers,
I am currently working with genomes from historical fungal samples and using SPAdes, along with other tools, for assembly.
I have a few questions about how best to run SPAdes in this context. As you can imagine, the DNA is degraded, fragmented, and coverage is uneven.
I have tried assembling the genomes using both paired-end Illumina reads and merged reads (in separate runs). Based on BUSCO scores, the differences are minor, and neither input type consistently produces better results. Would you recommend using paired-end reads rather than merged reads, or does this not make much difference for the algorithm?
I am also curious about how SPAdes handles reads shorter than the smallest k-mer in the list. Are these reads discarded? Do you have any recommendations for selecting k-mers? So far, I have tested both the "auto" mode (21, 33, 55) and a longer list (21, 27, 29, 31, 33, 35, 37, 41, 51, 61, 71, 81, 91, 101, 111, 121). The BUSCO results are very similar in most cases, so I wonder whether including a more extensive set of k-mers provides any practical benefit.
Finally, regarding the --isolate flag: the log suggests using it, but I have been hesitant given the probability of contamination in these historical samples. Do you have any recommendations on whether to enable this flag in such cases?
Thank you very much for your time and advice.
Best wishes,
Lia
Dear SPAdes developers,
I am currently working with genomes from historical fungal samples and using SPAdes, along with other tools, for assembly.
I have a few questions about how best to run SPAdes in this context. As you can imagine, the DNA is degraded, fragmented, and coverage is uneven.
I have tried assembling the genomes using both paired-end Illumina reads and merged reads (in separate runs). Based on BUSCO scores, the differences are minor, and neither input type consistently produces better results. Would you recommend using paired-end reads rather than merged reads, or does this not make much difference for the algorithm?
I am also curious about how SPAdes handles reads shorter than the smallest k-mer in the list. Are these reads discarded? Do you have any recommendations for selecting k-mers? So far, I have tested both the "auto" mode (21, 33, 55) and a longer list (21, 27, 29, 31, 33, 35, 37, 41, 51, 61, 71, 81, 91, 101, 111, 121). The BUSCO results are very similar in most cases, so I wonder whether including a more extensive set of k-mers provides any practical benefit.
Finally, regarding the
--isolateflag: the log suggests using it, but I have been hesitant given the probability of contamination in these historical samples. Do you have any recommendations on whether to enable this flag in such cases?Thank you very much for your time and advice.
Best wishes,
Lia