I have only shotgun sequences of metagenomes obtained without PCR of 16S rDNA genes. If there is a method to analyze taxonomic abundance from these shotgun sequences, please let me know some references.
Hi Hae-In Lee,
It is absolutely possible! In fact if your sequencing was of sufficient depth, it is better than 16S amplicon based taxonomic surveys. This is because it bypasses the biases due to PCR of the gene, you can read more in Kennedy et al (2014).
If you want to assess the coverage of your sequencing, you can use the tool provided by Rodrigeuz-r et al. (2014). It uses redundancy to assess coverage of your metagenome, which is very helpful in deciding which analyses to do downstream.
So, tools (a pipeline) I would use:
- Perform all the quality control on your dataset
- Perform a blast like similarity search of all your reads to a database, such as the NCBI non-redundant database. (The tool I use for this is DIAMOND, which is optimized for short metagenomic sequences)
- Next I would use MEGAN, which is very nice for taxonomic profiling, as well as functional profiling. MEGAN employs a graphic user interface and can create barcharts and many other cool plots out of you taxonomic data, and provides file which can be incorporated into programs such as STAMP, for statistical analysis.
Hope this helps. I prefer using raw reads for taxonomic profiling, as this allows you to assess the entire diversity sampled. Assembled reads and predicted genes are then ideal for functional characterization.
Kind regards,
Andries
Article Nonpareil: a redundancy-based approach to assess the level o...
Article Evaluating Bias of Illumina-Based Bacterial 16S rRNA Gene Profiles
Article STAMP: Statistical analysis of taxonomic and functional profiles
Article MEGAN Community Edition - Interactive Exploration and Analys...
Article Fast and sensitive protein alignment using DIAMOND
Awesome! Thank you very much, Andries. It was really helpful. By the way, would you provide more details of the references?
One way is also to upload your data to a database, such as MG-RAST, one-codex (very fast in producing results) or the IMG database (just google them and sign up). These database will do the work in classifying your sequences and generating graphs that you can explore. I ussually do it to complement my own analysis, with tools like MEGAN, or better kraken (command line)
Hi Hae-In Lee,
Do you still want more information on the references?
Nonpareil lets you estimate metagenome coverage.
STAMP is for statistical analysis on your metagenomes.
MEGAN is for analysing the taxonomy and function of your metagenome.
And DIAMOND is for assigning the taxonomy in the first place.
Hope this helps.
Andries
Hi Andries,
Your answer and the references you provided are really helpful. I think that is enough for now. DIAMOND and MEGAN are what I've used for functional analysis. I am learning myself how to use STAMP. It looks pretty awesome.
Thank you again for your help.
Hae-In
Hae-In Lee Andries Johannes Van der Walt Is it possible to screen the antimicrobial resistance genes from the shot-gun sequence data?
- Badges
- Science topic
- Similar topics
- Bioinformatics
- Bioinformatics and Computational Biology