Skip to content

Latest commit

 

History

History
136 lines (81 loc) · 4.12 KB

report.md

File metadata and controls

136 lines (81 loc) · 4.12 KB
Raw

Workflow Analysis

Why is this analysis necessary?

  • This workflow aims to identify diversity of bacteria species present in samples.

Application.

We used these two pipelines to identify the efficiency of different treatments (B,G and K) on bacteria species present in 15 dog samples, each with paired reads.

Objectives.

  1. To develop a 16S-rRNA analysis workflow from illumina sequencing platform.
  2. To test the pipelines.
  3. To analyse the results.
  4. To compare the tools used.

Methodology.

Quality Control.

Purpose - Quality Control is performed to inspect the quality of sequenced reads. The quality scores will guide in trimming and filtering.

Tools used -Fastqc ,Multiqc and QIIME2. Time - 2 minutes for both.

Pros - The tools are fast,lightweight and simple syntax. Cons- None

Conclusion: Not much of the reads were truncated because the quality of reads was good.

Trimming and Filtering reads.

Purpose - To remove poor quality reads.

Tools used -Trimmomatic, Prinseq and QIIME2

Time Taken:Trimmomatic - 10 minutes

       Prinseq - 6 minutes.
       Qiime - 4 minutes

Pros: - Trimmomatic and Prinseq provide more trimming options.

  - Qiime tool does both the quality check, filtering and trimming in one step ; syntax is brief.

Cons: For Trimmomatic and Prinseq, the proccess and report generation is not automatic.

Chimera Detection.

Purpose - To remove chimeras which can lead to spurious results.

Tools used -UCHIME, Qiime Deblur.

Time Taken:UCHIME - 70 minutes. Deblur - 6 minutes.

Pros - Deblur is fast and does chimera detection, denoising and feature table creation at once.Moreover, there is an informative report generated.

Cons - UCHIME requires a reference database.

OTU Picking.

Purpose - To find representative sequences/Sequence Variants for the samples. This is used when assigning taxonomy.

Tools used - USEARCH, Qiime Deblur.

Time Taken:USEARCH - 50 minutes. Deblur - 4 minutes.

Usearch uses closed-reference OTU picking whereas Qiime uses De novo OTU picking.
Qiime picked 209 OTUS whereas USEARCH picked 113 OTUS.

Pros - Deblur is fast, uses De novo OTU picking which ensures no sequences are disregarded. - Deblur allows you to assign taxonomy upto subspecies level.

Cons - USEARCH discards sequences that do not have a match in the reference database. - The OTUS picked in USEARCH are limited to the database used.

ASV Prediction.

Purpose - To find representative sequences/Sequence Variants for the samples. This is used when assigning taxonomy.

Tools used - Qiime Dada2. Time Taken: - 60 minutes.

282 sequence variants picked.

Pros - More Accurate.

Cons - USEARCH discards sequences that do not have a match in the reference database. - The OTUS picked in USEARCH are limited to the database used.

Classification and Alignment.

Purpose - To establish the similarities and dissimilarities between the variants picked (OTUS).

Tools used - Qiime mafft and FastTree. Time Taken:Qiime - 10 minutes.

Pros - Both Mafft and FastTree are fast and lightweight.

Cons - The phylogeny tree . - The OTUS picked in USEARCH are limited to the database used.

Diversity.
Alpha Diversity.

Purpose - To establish the similarities within a community or treatment.

Tools used - Qiime. Time Taken:Qiime - 7 minutes.

Pros - The core metrics analyses generates alpha and beta diversity results in one command.

Beta Diversity.

Purpose - To establish the similarities within a community or treatment. Tools used - Qiime. Time Taken:Qiime - 7 minutes.

Pros - The core metrics analyses generates alpha and beta diversity results in one command.

Results.

To view the results generated from this project, click here

Conclusion.

The results from both pipelines did not show significant differences on the dataset used. However, a bigger dataset would inform more about the most efficent pipeline.

There was a more diverse population found in samples treated with B compared to K and G(least diverse).