heterogeneity spacer in mapping file

[anke.s...@gmail.com]

Hi,

I am trying LotuS for the first time and wonder if my mapping file is correct.

I have an 18S amplicon library, generated with with the dual-indexing approach described by Fadrosh et al. 2014, Microbiome. I.e. I generated the amplicons with primers that each included a barcode sequence, a heterogeneity spacer and the actual primer sequence.
I have run LotuS using a mapping file where I just gave the barcode and primer sequences, and a I have run it with a file giving it it the barcodes and adding the heterogeneity spacers to the primers, see below. The results are quite different for some samples, so I wonder which approach I should use?
I allowed for one error in the barcode and two errors in the primer sequence, none of the barcodes or primers were reverse complemented (that is ok, isn't it), and I provided two fastq files as input files (read1 and read2 of a paired-end miSeq run)

Example without heterogeneity spacers

#SampleID	BarcodeSequence	Barcode2ndPair	ForwardPrimer	ReversePrimer
01AW	CCTAAACTACGG	TGTTGCGTTTCT	GTACACACCGCCCGTC	CCTTCYGCAGGTTCACCTAC
01BW	GTGGTATGGGAG	GTGGTATGGGAG	GTACACACCGCCCGTC	CCTTCYGCAGGTTCACCTAC
01KB	ATCTAGTGGCAA	GTGGTATGGGAG	GTACACACCGCCCGTC	CCTTCYGCAGGTTCACCTAC
02KB	TACCGCCTCGGA	ACAGCCACCCAT	GTACACACCGCCCGTC	CCTTCYGCAGGTTCACCTAC
02S	GTGGTATGGGAG	GAGCAACATCCT	GTACACACCGCCCGTC	CCTTCYGCAGGTTCACCTAC
02W	TACCGGCTTGCA	GAGCAACATCCT	GTACACACCGCCCGTC	CCTTCYGCAGGTTCACCTAC
03KB	TACCGCCTCGGA	GTGGTATGGGAG	GTACACACCGCCCGTC	CCTTCYGCAGGTTCACCTAC
03S	GTTACGTGGTTG	GAGCAACATCCT	GTACACACCGCCCGTC	CCTTCYGCAGGTTCACCTAC

Example with heterogeneity spacers

#SampleID	BarcodeSequence	Barcode2ndPair	ForwardPrimer	ReversePrimer
01AW	CCTAAACTACGG	TGTTGCGTTTCT	GTACACACCGCCCGTC	GACCTTCYGCAGGTTCACCTAC
01BW	GTGGTATGGGAG	GTGGTATGGGAG	GGTACACACCGCCCGTC	ACCTTCYGCAGGTTCACCTAC
01KB	ATCTAGTGGCAA	GTGGTATGGGAG	CACTGGTACACACCGCCCGTC	ACCTTCYGCAGGTTCACCTAC
02KB	TACCGCCTCGGA	ACAGCCACCCAT	ACTGGTACACACCGCCCGTC	AGACCTTCYGCAGGTTCACCTAC
02S	GTGGTATGGGAG	GAGCAACATCCT	GGTACACACCGCCCGTC	ACCTTCYGCAGGTTCACCTAC
02W	TACCGGCTTGCA	GAGCAACATCCT	CACTGGTACACACCGCCCGTC	ACCTTCYGCAGGTTCACCTAC
03KB	TACCGCCTCGGA	GTGGTATGGGAG	ACTGGTACACACCGCCCGTC	ACCTTCYGCAGGTTCACCTAC
03S	GTTACGTGGTTG	GAGCAACATCCT	ACTGGTACACACCGCCCGTC	ACCTTCYGCAGGTTCACCTAC

Kind regards, Anke.

[Falk Hildebrand]

Hey Anke,

also here sorry for the late answer, I just found these questions by setting up a new group and saw that email alerts were deactivated for this group. 

About the heterogenity spacers: Don't include the heterogenity spacer sequence in the fwd/ rev primer or the fwd/ rev primer. Please ensure that the Fwd/ Rev primer are always the same string, unless it's different sequencing runs. And even then, I do not recommend mixing metabarcoding experiments, were the LSU / SSU was amplified by different PCR primers.

As to reversing the primer: I have some automatic checks in the pipeline if the primers were reversed, this should not make a difference. In general I would also allow for 0 errors in the Barcode, unless the gain in identified sequences (e.g. more than ~2%) warrants the use of this option. But usually the gain is just not worth the possible loss in precision (LotuS does check though, if with 1 BC error, several sample assignments would be possible, and then excludes these cases). Actually the rev primer is not checked for being reverse complimented. Please ensure that you find the rev primer sequence as you have it in the mapping file like this in the start of read2 fastq. E.g. search in the file for the "ACCTTCYGCAGGTTCACCTAC" sequence. If you inserted it in the correct orientation, it should be in the 5' of the sequence (jsut at the start).

hth,
Falk