Commit 569a0140 authored by Mathieu Giraud's avatar Mathieu Giraud
Browse files

doc/ more software agnocism, highlight section 'what is a clone ?'

parent 862f7855
#+TITLE: .analysis and .vidjil format
#+AUTHOR: The Vidjil team
The =.analysis= and the =.vidjil= files share a common [[][.json]] format.
They are produced and used by several components of the Vidjil platform,
but you can also use these formats to use the Vidjil browser within
your own analysis pipeline.
The following [[][.json]] format allows to
encode a set of clones with V(D)J immune recombinations,
possibly with user annotations.
In Vidjil, this format is used by both the =.analysis= and the =.vidjil= files.
The =.vidjil= file represents the actual data on clones (and that can
reach megabytes). It should be automatically produced.
reach megabytes, or even more), usually produced by processing reads by some RepSeq software.
(for example with detailed information on the 100 or 1000 top clones).
The =.analysis= file describes customizations done by the user
(or by some automatic pre-processing) on the Vidjil browser. The browser
can load or save such files (and possibly from/to the patient database).
......@@ -16,6 +16,21 @@ It is intended to be very small (a few kilobytes).
All settings in the =.analysis= file override the settings that could be
present in the =.vidjil= file.
* What is a clone ?
There are several definitions of what may be a clonotype,
depending on different RepSeq software or studies.
This format and the Vidjil browser both accept any kind of definition:
Clones are identified by a =id= string that may be an arbitrary identifier such as =clone-072a=.
Software computing clones may choose some relevant identifiers:
- =CARPRDWNTYYYYGMDVW IGHV3-11*00 IGHJ6*00=, a CDR3 AA sequence with additional V/J gene information (MiXCR)
- the 'clone sequence' as computed by the ARReST in =.clntab= files (processed by
- see also 'IMGT clonotype (AA) or (nt)'
* Examples
** =.vidjil= file -- one sample
......@@ -68,10 +83,10 @@ Note that other elements could be added by some program (such as =tag= or =clust
** =.vidjil= file -- several samples
This a =.vidjil= file obtained by merging with two =.vidjil= files corresponding to two samples.
Clones that have a same =id= are gathered.
Clones that have a same =id= are gathered (see 'What is a clone?', above).
It is the responsability of the program generating the initial =.vidjil= files to choose these =id= to
do a correct gathering ('windows' is used by Vidjil, 'clone sequence' is used by EC-NGS/Brno pipeline,
and 'IMGT clonotype (AA) or (nt)' could also be used by some programs).
do a correct gathering.
#+BEGIN_SRC js :tangle analysis-example2.vidjil
......@@ -258,9 +273,7 @@ In the .analysis file, this section is intended to describe some specific clones
"id": "", // clone identifier, must be unique [required]
// Vidjil/algo output -> the 'window'
// Brno .clntab -> clone sequence
"id": "", // clone identifier, must be unique [required] [see above, 'What is a clone ?']
// the clone identifier in the .vidjil file and in .analysis file must match
"germline": "" // [required for .vidjil]
......@@ -358,21 +371,5 @@ The default tag names are defined in [[../browser/js/vidjil-style.js]].
"key" : "value" // "key" is the tag id from 0 to 7 and "value" is the custom tag name attributed
* Differences between programs
Due to specificities between programs, some elements may differ depending
on which program has been run.
** MiXCR
The output when using MiXCR differs from Vidjil on the id of each clone.
Where Vidjil provides the representative sequence of the clone, MiXCR
provides the representative sequence in =Amino Acids= followed by the name
of the =V gene= and the name of the =J gene=.
"germline": ...
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment