Tool:AmphoraNet, AmphoraVizu - metagenomic and genomic analysis online tools using phylogenetic marker genes

Question

[Csaba Kerepesi]

Let me to introduce two new metagenomics analysis tools:

AmphoraNet is the webserver implementation of the AMPHORA2 software which is a taxonomic composition analysis tool for metagenomic shotgun sequencing data. Using phylogenetic marker genes AmphoraNet is suitable for estimating the taxonomic composition of bacterial and archaeal communities faster and easier than other tools. Published in Gene.

AmphoraVizu is an online tool for visualizing outputs generated by the metagenomic analysis tool AMPHORA2 or its webserver implementation AmphoraNet. Published in Microbial Ecology

I welcome any comment, question and suggestion! 

 

Here is a printscreen of an output of AmphoraVizu run on a Chinese Type 2 Diabetes metagenome:

[marina.v.yurieva]

Does it show taxonomy only down to class or any lower levels?

[Csaba Kerepesi]

It can show any level.

AmphoraNet calculates any taxonomy levels and then using AmphoraVizu you can choose among Species, Genus, Family, Order, Phylum, Superkingdom levels (simple switch Lowest taxonomic rank on the AmphoraVizu page).

[Csaba Kerepesi]

I changed the sample pictures.

[marina.v.yurieva]

Will try it! Thank you!

Answer 1

[Csaba Kerepesi]

Some important notes for users of AmphoraNet and AmphoraVizu:

1. Citations show AmphoraNet are applied more times for genomic analyses Nature communications, Genome Announc. than for metagenomic analyses. Unfortunately I did not highlight this application possibility on the websites of the web servers. I have made up the shortage.
2. There is 50MB upload limit now, so if you want to analyize a larger metagenome you should split it. There are various methods to split large multiple fasta files. Examples:
   • head, tail commands in Linux.
   • with Biopython: Split large fasta file
   • Form topic on biostars: Question: How To Split A Multiple Fasta
3. Be careful when you use "Archaea and Bacteria" option of AmphoraNet, because the more archaeal marker protein (104 vs 31) causes overestimating of relative abundance of archaea.
4. Estimate relative abundance carefully when your results have few marker protein hits. My tests on simulated metagenomes show it should be at least 100 marker protein hits per taxon for accurate relative abundance estimation. I am working on the exact statistical analysis on this interesting problem.
5. Marker protein databases of AMPHORA2 are built in October, 2011 (according the hmm files: dnaG.hmm using 1324 bacterial genomes and 114 archaeal genomes. At this moment there are 3272 bacterial genomes and 180 archaeal genomes (according to EBI complete genomes).

I welcome any question and comment.