I recently used Illumina BaseSpace's 16S Metagenomics App to gather some preliminary data. After the analysis, there was identification of a number of "unknown species". I was wondering if there was anyway of seeing what the actual sequences are for these "unknown species". Illumina told me they didn't have those kinds of output files.
As a bit of background, I used the Illumina MiSeq NGS platform with paired end reads and used general primers for the V3-V4 region and adapters for the Illumina Nextera Indexing Kit. The project goal is to build a phlyogenetic representation of bacterial species found in the water of a number of caves that are geographically near each other. We want to see how closely these "unknown species" that the app identified relate to the known species in the samples.
This is a side project and I am not a bioinformaticist, so this is all very new to me.
Hi.
The first you have to check if that "unknown species" are reads of Enterobacteria phage PhiX174. This phage sequences are used to check the quality of the Miseq run.
You have to remove this sequences. I think that the easier way to do this is make a clustering of your sequences, take the reference sequence of each OTUs and make the taxonomical assignment using nt database (to find and delete the phage OTUs). After that you can make a phylogenetic tree with this reference sequences, that have been taxonomical assigned previously, and check where are the "unknown species" in this phylogenetic tree.
The software you can use are MOTHUR for the sequences filtering and clustering, BLAST for the taxonomical assignment, and I don't know if you can also use MOTHUR for make phylogenetic tree or another one.
Good luck with your data analysis!!!
You can simply BLAST the sequences against NCBI's 16s reference database. PhiX sequences will have a very poor match. If you want to remove PhiX first, map the reads against PhiX using bowtie, bwa, or some other mapping program (or use BLAST again). If Illumina doesn't provide the sequences that correspond to "unknown species" (or any other taxonomic assignment), it's a lousy program and you shouldn't use it. It should not be a black box into which you feed sequence data and accept the taxonomic output without being able to check assignments independently.
Hi,
The regions selected for identification "V3-V4" from 16S gene are a standard for analysis of bacterial communities, however identifying the species level is unreliable using only these regions. There are studies that indicate that the 16S gene only provides reliable results to the level of genre, sometimes not even that.
Try to use RDP database identification tools or use the blast against some database like GreenGenes.
- Badges
- Science method