To be edited:

This repository contains code and example data for LC-MS/MS and amplicon sequencing data integration and analysis of the CCE1706 dataset.

---

List of Notebooks provided in this repository and its use:

• ASV_raw_data_processing.ipynb
  
  • The ASV 16S and 18SV9 files are used here. The sample names are renamed similar to that of the mzxml file names. Also, the taxonomic information for each feature is combined with the OTU table.
• Elemental_composition.ipynb
  
  • The summary files (.tsv format) from SIRIUS and CANOPUS outputs are used to get the elemental composition and some elemental ratios for each feature. Also van-Krevelen plots can be made.
• DataCleanup_Preliminary_Analysis.ipynb of feature table (combine SIRIUS, Canopus info) and PCoA & PermANOVA
  
  • Perform preliminary data-cleaning steps on feature table from MZmine. Also obtain PCoA plots on both Metabolites (feature-table) data and ASV data
• StackedBarPlot_ASV.ipynb
  
  • Using the ASV table (16S and 18SV9) and metadata, different stacked bar plots can be made for a particular taxonomic level binning (Phylum level or Genus level) according to different cycles (Cycle 1 Day 1 --> Cycle 4 Day 2)
• StackedBarPlot_Metabolites.ipynb
  
  • Using the feature table, different stacked bar plots can be made for a SuperClass level binning (Superclass as given by SIRIUS) according to different cycles . In addition to that, the binned information are saved and will be used further for Sunburst visualization
• Sunbursts_CCE_Metabolites_python.ipynb
  
  • Sunburst visualisation (in Python Notebook) - for Metabolites data ( sunburst plots showing levels: Superclass, Class, Subclass, Level 5)
• Sunburst visualisation (in Python Notebook) - until Genus level for sequencing data
  
  • Sunburst visualisation for ASV16s and ASV18sv9 data for 3 different depth ranges (full depth, surface:0-20m, deep:20-100m)
  • sunburst plots showing levels for asv18s: "Domain", "Kingdom", "Phylum", "Class", "Order", "Family", "Genus"
  • sunburst plots showing levels for asv16s:"Domain", "Phylum", "Class", "Order", "Family", "Genus", "Species"
• Correlation Analysis script: (Focus on explanation of FDR, how you do it and justification for the method)
  
  • to be updated
  • already available in the homepage. Needs to be commented better
• To list the important features, we tried RF and XGBoost (But we should focus on surface samples with different cycles)
  
  • to be updated
• Random_Forest.ipynb
  
  • To list the important features that helps to classify the data according to different cycles, random forest was performed.

FBMN links:

Analogs OFF, Analogs ON