Upload your OTU/ASV table
[For detailed information on how the files have to look, check out the Info & Settings tab on the left!]
Upload your MSD data directly into Namco
You can either upload a file with mulitple links from MSD and then choose one or paste a single link directly into the corresponding field.
Upload fastq sequencing files
[For detailed information on file specifications, check out the Info & Settings tab on the left!]
For details on the sample-name cutoff feature, check out the sample-names explanation in the 'Info & Settings' tab!
Select one of the following two amplicon sequencing analysis pipelines to process the fastq files:
Primer trimming in DADA2 is position based, so you need to know the length of your used primers. It is also assumed that primer sequences are at the beginning (left) of each read. If primers are already removed from your reads, enter 0.
The quality profiles above can guide you to find more fitting cutoff values
Primer trimming in LotuS2 is sequence based, so you need to know the sequence of your used primers. The tool will search for this sequence in each read and remove it.
Lotus2 will build a phylogenetic tree by default. If you do not need it, simply insert -buildPhylo 0 in the text field above.
Upload data from MSD
Explain MSD + link
Start by uploading your data or use our provided sample dataset and try out all the features in NAMCO
Documentation
Issues & Recommendations
News
Authors
References
Overview of NAMCO
Data Overview & Filtering
Please consider filtering your data!
Explore the meta-file you uploaded:
Here you can filter samples and taxonomic levels
Filter options for samples
Select which sample you want to remove; you can select samples by name or remove whole sample groups.
Filter options for taxa
Select which taxonomic features you want to keep (you can select multiple).
Filter History
Apply more advanced filterings onto your dataset. Click the checkbox to the left of a filtering function and select a fitting value; multiple combinations of functions can be applied. Click 'Apply Filter' when you are done.
The plots to the left show you the current distribution of the corresponding value as well as a red line to indicate the currently selected filtering value.
Note: these functions are applied to the normalized dataset!
Remove all OTUs with a summed up abundance value over all samples below ... (left of red line gets removed)
Remove all OTUs with a relative abundance value in all samples below ... %
The red line shows the chosen cutoff in relation to all appearing rel. abundance values.
Remove all OTUs which occur at least ... times in all samples (meaning, they have at most ... times an abundance value of 0) (right of red line gets removed)
Keep only those ... OTUs with the highest variance in abundance (left of red line gets removed)
Keep only those OTUs whith a prevalence value over ... %
Remove contaminated OTUs/ASVs using DNA concentration and/or control samples.
Tab-Information
Step 1: Select corresponding data columns
(a) DNA concentration
(b) Prevalence
Library size per sample
The above plot shows the library size of each sample. Control samples are usually located at the lower end of the library size distribution in a dataset.
Download as PDFScroll down once finished to inspect the detected contaminants...
Step 2: Inspect possible contaminants
Step 3: select and remove contaminants
Feature frequency vs DNA concentration
The above plot shows if the selected feature (OTU/ASV) is a robust candidate for a contaminant. The red line shows the model of a contaminant sequence feature, for which frequency is expected to be inversely proportional to input DNA concentration. This means if the black points follow this red line, this feature is in line with the model and can be considered a robust result.
Download as PDFPrevalence of features in control vs true samples
The above plot shows how often the individual OTUs/ASVs were observed in the control and true samples. Contaminants should clearly form a branch that distinguishes them from features that are not contaminated. You can hover over the individual points to find out the name of potential outliers.
Download as PDFUpload a meta-file, which assigns groups and values to your samples.
fastq Overview
Analysis of sequence quality for provided fastq files after pipeline
foreward
reverse
Number of reads after each step in DADA2 pipeline
Access logging infos from chosen pipeline
[Currently only available for LotuS2 pipeline]
Demultiplexing log
Complete run log
Download options
Download the generated ASV-tables:
Download the ASV sequences:
Download the taxonomic classification:
Download a phyloseq R-object:
Basic Analysis
Analyse samples by their taxonomic composition
Tab-Information
Options
Change order of y axis
Analyse the diversity of species inside of samples
Tab-Information
Raw values for alpha diversity scores, including download:
Download TableAnalyse the diversity of species between samples using non-euclidean distance measures
Tab-Information
Hierarchical clustering (Ward's method) of the sample using the chosen distance method
Download as PDFDimensionality reduction based on ecological distances
Change position of text in scatterplot
Analysis of species richness with rarefaction curves
Tab-Information
Differential Analysis
Explore different measures of association between sample groups
Tab-Information
Explore OTU topics in your dataset
Tab-Information
Input Variables:
To download as PNG, klick the photo-icon in the plot. PDF downloas is not supported yet.
Topics colored red, have a strong association with the chosen reference level; the blue topics on the other hand are associated with the other level within the chosen covariate. (Example: Chosen covariate is Gender and reference level is Female -> Female will be colored red, Male is blue)
To download as PNG, klick the photo-icon in the plot. PDF download is not supported yet.
Compare time-points and find trends in your dataset
Tab-Information
In order to start your analysis, select the correct meta groups in the options menu & press the 'Perform Analysis' button
Here you can see the significance value for each taxa and meta variable regarding the selected time-points. Since the values are -log10 transformed, the ones at the top behave significantly different over the time-points.
Download as tableOptions
Fixed Options
If you change a variable here, you need to hit the 'Perform Analysis' button again.
Interactive Options
If you chose to cluster your samples, you can find additional statistics & information down here:
(You first have to select how many clusters you want in the 'Options' menu above)
(The coloring is controlled by the groups over time-points selection above)
See, which sample was selected into which cluster:
Use biomehorizon package for time series visualization:
Tab-Information
Horizon options
Mandatory options
Further specifications
Perform statistical tests on different taxonomic levels between sample groups
Tab-Information
Select a multiple testing correction method:
Options
Functional Analysis
Functional prediction using Picrust2
Tab-Information
Picrust2 parameters & Input Files
Note: Picrust2 will always be applied on the non-normalized dataset automatically.
Next to the functional assignment of OTUs, Picrust2 also infers the copy numbers of each 16s-rRNA gene per OTU; you have the option to normalize your abundance values with the copy-numbers by selecting this checkbox.
Download zip-archive with raw picrust2 results:
Please be aware:
This will create a zip archive of all output files, so it might take a few seconds until the download window appears!
This download window will not appear if you use a restored dataset!
Differential analysis parameters
A higher number of MC iterations will increase precision of estimating the sampling error but also increase runtime. For datasets with few samples a higher value can be chosen, with more samples a lower one should be used.
This button is only activated if you have run the picrust analysis!
You can change the parameters on the left and rerun the analysis & reload the plots by pressing this button.
Download results of analysis (EC) Download results of analysis (KO) Download results of analysis (PW)Differential functional analysis
- EC
- KO
- PW
- Information & Options
- Downloads
- Relationships between effect size & p-value
Names of significant functions
Details about significant functions:
Here the functions with BH adjusted P-value above the significance threshold are displayed; the boxplot shows the different abundance distributions of a function colored by each sample group. Also the BH-adjusted P-value and effect size is displayed as a barplot.
Names of significant functions
Details about significant functions:
Here the functions with BH adjusted P-value above the significance threshold are displayed; the boxplot shows the different abundance distributions of a function colored by each sample group. Also the BH-adjusted P-value and effect size is displayed as a barplot.
Names of significant functions
Details about significant functions:
Here the functions with BH adjusted P-value above the significance threshold are displayed; the boxplot shows the different abundance distributions of a function colored by each sample group. Also the BH-adjusted P-value and effect size is displayed as a barplot.
Options for Visualization
Here you can set the significance cutoff for the BH adjusted P-value; functions with a p-value below it are considered significant.
Here you can set the significance cutoff for the effect size; functions with a effect size greater than it it are considered significant.
If too many points in close proximity are considered significant, change the number of overlaps, to display more labels.
Information
Phylogenetic Analysis
Phylogenetic Tree of OTU taxa
Tab-Information
Basic tree visualization options:
Advanced tree visualization options:
Additional options to manage tree:
Network Analysis
Co-occurrences network generation
Tab-Information
Information
You can look at the following two plots to see the effect of your chosen cutoff:
Cutoff-Barplot
Cutoff-Heatmap
Information
Options
chosen parameters
Green edges: OTU-pair more often occuring in selected reference group.
Red edges: OTU-pair more often occuring in other sample group, which is compared against.
Create a single network on your whole dataset
Tab-Information
Parameter-information
Additional Parameters
Display options
Node options:
Edge options:
Other options:
Details about network
Explore network structures between the discovered taxonomic ranks
Tab-Information
Parameter-information
Additional Parameters
Display options
Node options:
Edge options:
Other options:
Details about network
Explore network structures in different sample groups
Tab-Information
Parameter-information
Additional Parameters
Display options
Node options:
Edge options:
Other options:
Details about network difference
Confounding Analysis
Analyse confounding factors
Tab-Information
This heatmap will show you if there are confounding factors for specific variables. Read the plot from x to y axis like this: tested variable XX has possible confounding factors YY (if the legend says 'yes', YY is a confounder for XX).
Explained Variation:
The points represent the indiviual groups that are plotted over their negative log 10 p-value and r-square value.
Options
Machine Learning
Build a Random Forest machine learning model
Tab-Information
Options for building the model:
Model parameters:
Resampling options:
Results
Confusion Matrix for testing-dataset
Download as PDFConfusion Matrix for full dataset
Download as PDFROC-Plot: TP-rate vs. FP-rate including AUC for model
Download as PDFThe receiver operating characteristic (ROC) can show you how good the model can distuingish between sample-groups. A perfect ROC-Curve would go from (0,0) to (0,1) to (1,1). This means the model has a perfect measure of seperability. A ROC-Curve that goes diagonally from (0,0) to (1,1) tells you, that the model makes only random predictions.
The AUC (area under the curve) is a good measure to compare multiple ROC curves and therefore models. Here a AUC of 1 tells you, that you have a perfect model, AUC of 0.5 is again only random.
Show the top x most important features for building the model. You can change how many features to display by moving the slider.