no matter how modify it ,it will be like this ,subsequent cannot be classified correctly.
no matter how modify it ,it will be like this ,subsequent cannot be classified correctly.
Hello!
What is a length of the region you are targeting? It can be that your sequences after truncating are too short and do not overlap. There are several options that can help:
- Set a higher values for truncation
- Decrease min overlap region
Best,
1 Like
I analyzed two batches of different data together, and this batch of forward and reverse was 249 in length, so I tried 240 truncation
But the other batch of data is only 220 in length, and when I try to analyze it with 240, I will report an error
I was completely unable to combine these two batches of data for a unified analysis。When I analyze it together, there is always a batch of data that cannot be combined。
Truncation parameter should not be higher than the lengthe of the reads, since all reads with shorter length will be discarded.
If you want to analyze both batches, you need to process them separately by dada2, setting the same parameters, which satisfy both batches. So, truncation parameter should be around 215. If reads failed to merge, you need to:
1 Decrease min overlapp parameter
2 If option 1 do not work, you may consider to use only forward reads for the analysis.
1 Like
This is a visualization of the quality of the two batches of data I entered together，
After I adjusted the minimum value of the overlap to 4, when the cut length was all 220, the result of this batch of data of 249 was
only forward reads for the analysis，Does it mean using deblue to merge and then analyze?
Quality plot of forward reads looks strange to me - never saw a plot like this before. If others mods can clarify it - please jump in 
.
Using only forward reads do not mean to join reads and use deblur - it's mean that you can disregard reverse reads and process only forward reads by running Dada2 single instead of Dada2 paired. Taxonomy annotations will suffer from it since sequences will be shorter.
After the data was taken back from the company, I did fastqc, almost all the data quality is above Q20, it should be filtered in advance.
Using only forward reads ，Will Taxonomy annotations be greatly affected? Is there a lot of difference from normal bidirectional merges? I haven't tried that, I'll try。
Is it possible to generate two batches of data into their own abundance information tables and then use a script to merge them。
Yes, you can process two batches separately and after it merge both in one feature table ans one rep-seq.qza, but Dada2 parameters should be the same in both cases.
Taxonomy annotations definitely will be affected, but I can not say at which extent since it may depend on the region you targeted.
I would try to process only forward reads with the same parameters for both batches, merge output files and annotate them to see if taxonomy annotations are good enough to keep this approach.
This topic was automatically closed 31 days after the last reply. New replies are no longer allowed.