mapping file and add qiime labels

[Guðný Vala Þorsteinsdóttir]

Hi, I'm new to qiime and I have some questions about how to make a mapping file and adding labels to my samples.

I have paired-end reads from Illumina Miseq and my samples are already demultiplexed so I have no barcodes but linker primer on both ends. The reads are in fasta files per sample (8 files). So my question is, how do I make this mapping file and add IDs to my samples to combine the data? I have 8 samples in total, but I want to analyze them both at 8 sample basis and also by two groups (4 samples per group). Do I only make one mapping file or do I have to make another depending on how I want to analyze by groups/samples? 

thanks!

GV

[Justine Debelius]

First, I strongly recommend combining your samples for the annotation step. If you're including any kind of a de novo step, you cannot compare reads which were not clustered together, since there is no comparison between the OTUs or sOTUs. My suggestion (although I know others disagree) is that if you're using OTU-based clustering, you perform a reference-based step first, if this is an option. My preference toward reference-based clustering is that it allows consistency in annotation across studies: if you find OTU X and I find OTU X, then we're probably seeing similar sequences, which may mean we're seeing similar organisms.

You will need to use join_paired_ends.py or multiple_join_paired_ends.py to concatenate your reads to concatenate your reads.

The names need to appear in the mapping file. I suggest putting everything in the same mapping file so you can run the commands once. Demultiplexing is done independently, so combining the files has no effect. You can then use this mapping file with multiple_split_libraries.py.

Best,

Justine