Pipelines

Overview

Pipeline

Species

Features and Repo

Author

Clockwork combined

TB + NTM

  • Speciation (Kraken2 + Mykrobe)

  • Clockwork (for TB)

  • Drug Resistance on Clockwork VCF

  • TB Neighbourhood and Phylogeny Trees

Denis & Fan

Clockwork VC

TB

https://github.com/iqbal-lab-org/clockwork

Martin Hunt

Oxford CompassCompact

TB + NTM

https://github.com/oxfordmmm/CompassCompact

Yifei & Fan

APHA BTB

Bovine TB

https://github.com/oxfordmmm/BovTB-nf-docker

Richard & Fan

Oxford Bug Flow

Bacteria

https://github.com/davideyre/bug-flow

David Eyre

TB Pipelines

pipleine

input

kraken2

removecontam

mykrobe

auto pick reference

trim

bwa map

samtool mpileup

samtoolqc

cortex

minos

fasta

resistance

catbug tb

fastq/bam

Y

Y

Y

Y

Y

Y

Y

Y

Y

Y

Y

Y

catbug non-tb

fastq/bam

Y

Y

Y

Y

Y

Y

Y

Y

Y

clockwork tb

fastq/bam

Y

Y

Y

Y

Y

Y

Y

Y

Y

Y

Y

Y

CompassCompact

fastq/bam

Y

Y

Y

Y

Clockwork VC

fastq

Y

Y

Y

Y

Y

Y

TB Neighbourhood (alpha release)

For clockwork TB pipeline, we have a service built for examining genetic neighbourhood based on SNP distance.

To see genetic related samples, a sample would have been through following processes:

  1. The sample has been succesfully run in clockwork pipeline.

  2. The sample has been identified by mykrobe as mycobacterium tuberculosis.

  3. The sample has been succesfully mapped to reference NC_000962.3.

  4. The sample has been archived to Sp3 persistence storage. (This is currently done manually.)

  5. The sample has been pushed to TB neighbourhood service, catwalk. (This is currently done manually.)

With alpha release, you would not see neighbourhood right after the run as the step 4 and step 5 would be run manually at the moment. The step 4 and step 5 would be done at the end of the day or end of the week, so please check it later.

If you have a sample that has been through all the above processes, here is how to see the genetic neighbours.

Step 1

Click the sample name to view SP3 report, from a complete run, eg. click ERR552831 as shown below.

_images/neighbour11.png

Step 2

Click link “Query genetic neighbourhood” at the top of the SP3 report below the sample name.

_images/neighbour22.png

Step 3 (with new features)

Genetic neighbourhood page would show you the result. In the example below, there were 2 neighbours found shown in a neighbourhood table.

These two neighbours are from EBI Cats site (other sites samples could be detected if there are any).

The run name of the neighour samples (with a prefix of organisation) gives a hint on the source of the neighbours, in this case, oxforduni.

And the sample names of the neighbours are also shown at the 3rd column of the table.

The last column of the neightbour table is the SNP distance, to the query sample.

_images/neighbour33.png
  1. There will be no duplicate samples among neighbours. (although there might be one sample same name as the query sample)

  2. Quick select for samples from your own organisation. (e.g. in the example, select samples prefixed as “sp3test1”)

  3. Quick select for samples from other organisations. (e.g. in the sample, select samples prefixed other than “sp3test1”)

  4. You can always select and unselect all samples using the checkbox near “Site name”.

  5. You have to select 3 or more samples to build a tree.

You can also query genetic neighbours ad-hoc in an interactive way, go to “clockwork” pipeline and click “View TB Neighbourhood”

_images/viewTB.png

Follow the hint, use a run id and a sample name and set a SNP distance for the query.

_images/neighbour.png

TB Phylogeny

For TB Neighbourhood, we build phylogeny trees for the nearest neighbourhood of TB samples.

To create a tree for your sample and its neighbours:

Step 1. Query neighbours

At SP3 sample report page, click ‘Query genetic neighbourhood’.

_images/tree1.png

Step 2. Select and generate

Select the samples among the neighbours, and build the tree.

You can adjust SNP distance, and select all or part of the neighbours to build trees.

_images/tree2.png

Step 3. Confirm and name the tree

Confirm your selection, and provide a name for your tree for later reference. Then click the button to generate the tree.

_images/tree3.png

Step 4. Monitor the queue

The tree buildling could be compute intensive and it could take a while.

After tree building job is submitted, you would be able to see your tree job is queued to compute.

_images/tree4.png

You can refresh the page, until it completes. It could be seconds, or minutes or hours depending on the amount of the nodes.

_images/tree5.png

Step 5. View the tree

When completed, the tree can be viewed in a page. (This uses phylocanvas, so javascript needs to be allowed in the browser.)

_images/tree6.png

In this page, one or more nodes can be highlighted, given a search text in the node label.

_images/tree7.png

You can use more features of Phylocanvas, by right click on the tree, such as:

  • Export leave labels, newick file and tree image

  • Show or hide labels

  • Align/Realign labels

You can also zoom or pan the image or click on node to highlight or clear the highlight.