warnings.md 2.28 KB
Newer Older
1
# Warnings in `.vidjil` files
2 3 4 5 6 7 8 9 10 11

These warnings are found in `.vidjil` files and alert the user on some unusual result.
They will be displayed in the Vidjil web application. Some warnings are related to individual clones or reads,
whereas some others are related to the dataset or the programs used. They can be produced by vidjil-algo
or by any RepSeq program or script willing to inform the user.

The following list is a work-in-progress, with links to gitlab issues in Vidjil.
Warnings which were implemented ([x]) have a fixed code that should not be changed for backward compatibility.


12
## Generic errors
13 14 15 16
- [ ] W0x Outdated program
- [ ]     Outdated germlines  
- [ ]     Strange parameters: xxxx xxxx
- [ ] W0z Unknown error
17
- [x] W09 Program interrupted, output data may be not complete
18

19
## Output of a pre-process ~"server-pre-process" 
20 21 22 23
- [ ] W10 Few assembled reads  #2243  
- [ ] W1z Other pre-processing warning/error


24
## Output of an analysis, global warnings
25 26 27 28 29
- [x] W20 Very few V(D)J recombinations found (0.7%)
- [ ] W2x Sequences with known adapters #1669
- [ ] W2x CDR3 detection without gapped germlines   #2187   (ou bien par clone ?)


30
## Output of an analysis, warnings on some clones
31

32
### Read quality
33 34
- [ ] W40 Low quality  #1544 

35
### Clone quality (KmerSegmenter in vidjil-algo)
36 37 38 39 40 41
- [x] W50 Short or shifted window in vidjil-algo
- [x] W51 Low coverage (0.112)
- [ ]     Bad e-value (0.xxx)  #1437/#1566/#1889 
- [x] W53 Similar to another clone      
- [ ]     Possible strand ambiguity 

42
### Strange recombination (FineSegmenter in vidjil-algo), V and J genes
43
- [ ]     Potential co-linear genome match (pos xxxxxx) #1664 #1629 
44
- [ ] W61 Non-recombined D7-27/J1 sequence  #2232
45 46 47 48 49
- [ ]     Potential di-mer #2820 
- [ ]     Very large deletion (xxx bp) #2909
- [ ]     Unexpected recombination
- [ ]     High probability clone ?
- [ ] W6z Unexpected error in V(D)J analysis (FineSegmenter failed)
Mathieu Giraud's avatar
Mathieu Giraud committed
50
- [x] W69 Several V/J candidate genes with equal probability #3575
51

52
### Strange recombination (FineSegmenter in vidjil-algo), D gene, N regions*
53 54 55 56 57
- [ ] W7x Mutations near breakpoint #1412 
- [ ]     Palindromic sequence ?
- [ ]     D with bad e-value ? 


58
## Comparisons between several samples or patients (such as with clonedb)
59 60 61
- [ ] W8x Potential contamination or public clone #1744 
- [ ]     Known public clone