How to just align for constant region

[ShaowenJ]

Dear Amazing Mixcr developers,

You guys are very helpful for us on setting the previous Nanopore long-read preset. The results look fantastic, but there are only 25% of all the reads have aligned successfully (It's a target pull-down library specific for TCR/BCR). We are wondering what are the unaligned reads come from.
Is that a way I can just check if the reads can align to the constant region?

The preset I am using is: under version of v4.2.0

flags: []
pipeline:
- align
- assemble
- assembleContigs
- exportClones
align:
  species: mmu
  libraryName: default
  trimmingQualityThreshold: 0
  trimmingWindowSize: 0
  chains: ALL
  replaceWildcards: true
  overlapPairedReads: true
  bamDropNonVDJ: false
  writeFailedAlignments: false
  tagPattern: null
  tagUnstranded: false
  tagMaxBudget: 10.0
  sampleTable: null
  splitBySample: false
  inferSampleTable: false
  readIdAsCellTag: false
  limit: null
  parameters:
    vParameters:
      geneFeatureToAlign: VTranscriptWithP
      minSumScore: 150
      relativeMinScore: 0.97
      parameters:
        type: kaligner2
        mapperNValue: 14
        mapperKValue: 1
        floatingLeftBound: true
        floatingRightBound: true
        mapperAbsoluteMinClusterScore: 102
        mapperExtraClusterScore: -38
        mapperMatchScore: 148
        mapperMismatchScore: -14
        mapperOffsetShiftScore: -82
        mapperSlotCount: 12
        mapperMaxClusters: 8
        mapperMaxClusterIndels: 4
        mapperKMersPerPosition: 4
        mapperAbsoluteMinScore: 100
        mapperRelativeMinScore: 0.8
        mapperMinSeedsDistance: 5
        mapperMaxSeedsDistance: 10
        alignmentStopPenalty: 0
        absoluteMinScore: 194
        relativeMinScore: 0.73
        maxHits: 7
        scoring:
          type: affine
          alphabet: nucleotide
          subsMatrix: "simple(match = 10, mismatch = -20)"
          gapOpenPenalty: -62
          gapExtensionPenalty: -11
    dParameters:
      geneFeatureToAlign: DRegionWithP
      relativeMinScore: 0.85
      absoluteMinScore: 25.0
      maxHits: 3
      scoring:
        type: linear
        alphabet: nucleotide
        subsMatrix: "simple(match = 5, mismatch = -9)"
        gapPenalty: -12
    jParameters:
      geneFeatureToAlign: JRegionWithP
      minSumScore: 150
      relativeMinScore: 0.97
      parameters:
        type: kaligner2
        mapperNValue: 12
        mapperKValue: 1
        floatingLeftBound: true
        floatingRightBound: true
        mapperAbsoluteMinClusterScore: 102
        mapperExtraClusterScore: -38
        mapperMatchScore: 134
        mapperMismatchScore: -14
        mapperOffsetShiftScore: -82
        mapperSlotCount: 11
        mapperMaxClusters: 12
        mapperMaxClusterIndels: 4
        mapperKMersPerPosition: 4
        mapperAbsoluteMinScore: 100
        mapperRelativeMinScore: 0.94
        mapperMinSeedsDistance: 6
        mapperMaxSeedsDistance: 8
        alignmentStopPenalty: 0
        absoluteMinScore: 112
        relativeMinScore: 0.78
        maxHits: 3
        scoring:
          type: affine
          alphabet: nucleotide
          subsMatrix: "simple(match = 10, mismatch = -20)"
          gapOpenPenalty: -62
          gapExtensionPenalty: -11
    cParameters:
      geneFeatureToAlign: CExon1
      minSumScore: 40
      relativeMinScore: 0.97
      parameters:
        type: kaligner
        mapperKValue: 11
        floatingLeftBound: false
        floatingRightBound: false
        mapperAbsoluteMinScore: 1.5
        mapperRelativeMinScore: 0.75
        mapperMatchScore: 90.0
        mapperMismatchPenalty: -0.1
        mapperOffsetShiftPenalty: -0.3
        mapperMinSeedsDistance: 9
        mapperMaxSeedsDistance: 18
        minAlignmentLength: 15
        maxAdjacentIndels: 2
        alignmentStopPenalty: -1000
        absoluteMinScore: 22.0
        relativeMinScore: 0.75
        maxHits: 6
        scoring:
          type: linear
          alphabet: nucleotide
          subsMatrix: "simple(match = 5, mismatch = -9)"
          gapPenalty: -12
    vjAlignmentOrder: VThenJ
    libraryStructure: Unknown
    includeDScore: false
    includeCScore: false
    minSumScore: 120.0
    maxHits: 5
    relativeMinVFR3CDR3Score: 0.7
    allowPartialAlignments: false
    allowNoCDR3PartAlignments: false
    allowChimeras: false
    readsLayout: Opposite
    mergerParameters:
      qualityMergingAlgorithm: MaxSubtraction
      partsLayout: null
      minimalOverlap: 13
      minimalMatchQualitySum: 364
      maxQuality: 50
      minimalIdentity: 0.7
      identityType: MinimalQualityWeighted
    fixSeed: true
    alignmentBoundaryTolerance: 5
    minChimeraDetectionScore: 120
    vjOverlapWindow: 3
    saveOriginalSequence: false
    saveOriginalReads: false
    smartForceEdgeAlignments: true
refineTagsAndSort: null
assemblePartial: null
extend: null
assemble:
  sortBySequence: false
  clnaOutput: true
  cellLevel: false
  consensusAssemblerParameters: null
  cloneAssemblerParameters:
    assemblingFeatures:
    - CDR3
    minimalClonalSequenceLength: 12
    qualityAggregationType: Max
    cloneClusteringParameters:
      searchDepth: 2
      allowedMutationsInNRegions: 1
      searchParameters: twoMismatchesOrIndels
      clusteringFilter:
        type: relativeConcentration
        specificMutationProbability: 0.01
    cloneFactoryParameters:
      vParameters:
        relativeMinScore: null
        scoring: null
        maxAlignmentWidthLinear: 5
        maxAlignmentWidthAffine: 500
      jParameters:
        relativeMinScore: null
        scoring: null
        maxAlignmentWidthLinear: 5
        maxAlignmentWidthAffine: 500
      cParameters:
        relativeMinScore: null
        scoring: null
        maxAlignmentWidthLinear: 5
        maxAlignmentWidthAffine: 500
      dParameters:
        relativeMinScore: null
        absoluteMinScore: null
        maxHits: null
        scoring: null
      vdcParameters:
        Variable:
          relativeMinScore: null
          scoring: null
          maxAlignmentWidthLinear: 5
          maxAlignmentWidthAffine: 500
        Joining:
          relativeMinScore: null
          scoring: null
          maxAlignmentWidthLinear: 5
          maxAlignmentWidthAffine: 500
        Constant:
          relativeMinScore: null
          scoring: null
          maxAlignmentWidthLinear: 5
          maxAlignmentWidthAffine: 500
    separateByV: false
    separateByJ: false
    separateByC: false
    maximalPreClusteringRatio: 1.0
    preClusteringScoreFilteringRatio: 2.0
    preClusteringCountFilteringRatio: 2.0
    addReadsCountOnClustering: false
    badQualityThreshold: 0
    maxBadPointsPercent: 0.7
    mappingThreshold: 2of5
    minimalQuality: 15
    postFilters: null
  inferMinRecordsPerConsensus: false
assembleContigs:
  ignoreTags: false
  parameters:
    branchingMinimalQualityShare: 0.1
    branchingMinimalSumQuality: 60
    decisiveBranchingSumQualityThreshold: 120
    alignedSequenceEdgeDelta: 3
    alignmentEdgeRegionSize: 7
    minimalNonEdgePointsFraction: 0.25
    minimalMeanNormalizedQuality: 5.0
    outputMinimalQualityShare: 0.75
    outputMinimalSumQuality: 0
    subCloningRegions: null
    assemblingRegions: null
    postFiltering:
      type: NoFiltering
    trimmingParameters:
      averageQualityThreshold: 10.0
      windowSize: 8
    minimalContigLength: 20
    alignedRegionsOnly: false
exportAlignments:
  chains: ALL
  noHeader: false
  fields:
  - field: -targetSequences
  - field: -targetQualities
  - field: -vHitsWithScore
  - field: -dHitsWithScore
  - field: -jHitsWithScore
  - field: -cHitsWithScore
  - field: -vAlignments
  - field: -dAlignments
  - field: -jAlignments
  - field: -cAlignments
  - field: -allNFeaturesWithMinQuality
  - field: -allAAFeatures
  - field: -defaultAnchorPoints
exportClones:
  splitByTagType: null
  filterOutOfFrames: false
  filterStops: false
  chains: ALL
  noHeader: false
  fields:
  - field: -cloneId
  - field: -readCount
  - field: -readFraction
  - field: -targetSequences
  - field: -targetQualities
  - field: -vHitsWithScore
  - field: -dHitsWithScore
  - field: -jHitsWithScore
  - field: -cHitsWithScore
  - field: -vAlignments
  - field: -dAlignments
  - field: -jAlignments
  - field: -cAlignments
  - field: -allNFeaturesWithMinQuality
  - field: -allAAFeatures
  - field: -defaultAnchorPoints
  splitFilesBy:
  - geneLabel:ReliableChain
  groupClonesBy: []

Much appreciated!
Thanks again!

[mizraelson]

Hi,
Can you please share the exact command you run?

[ShaowenJ]

Of course,

mixcr analyze local:mixcr_MODIFY_ont_preset_param \
     -Xmx40g -f --verbose \
     --species mmu \
     --not-aligned-R1 ./Mixcr_demul_Infected/Infected_unaligned.fastq \
     -Malign.parameters.saveOriginalReads=true \
     input.fastq \
     ./Mixcr_demul_Infected \
     1> mixcr.log 2> mixcr.err

[mizraelson]

Thank you. And also the alignment report produced by mixcr will be helpful.

[ShaowenJ]

============== Align Report ==============
Analysis time: 24.28s
Total sequencing reads: 127641
Successfully aligned reads: 22834 (17.89%)
Alignment failed, no hits (not TCR/IG?): 73046 (57.23%)
Alignment failed because of absence of J hits: 31761 (24.88%)
Overlapped: 0 (0%)
Overlapped and aligned: 0 (0%)
Alignment-aided overlaps: 0 (NaN%)
Overlapped and not aligned: 0 (0%)
No CDR3 parts alignments, percent of successfully aligned: 59 (0.26%)
Partial aligned reads, percent of successfully aligned: 450 (1.97%)
TRA chains: 280 (1.23%)
TRA non-functional: 42 (15%)
TRB chains: 769 (3.37%)
TRB non-functional: 65 (8.45%)
TRD chains: 4 (0.02%)
TRD non-functional: 2 (50%)
TRG chains: 2 (0.01%)
TRG non-functional: 2 (100%)
IGH chains: 6203 (27.17%)
IGH non-functional: 392 (6.32%)
IGK chains: 11429 (50.05%)
IGK non-functional: 533 (4.66%)
IGL chains: 4147 (18.16%)
IGL non-functional: 999 (24.09%)
Realigned with forced non-floating bound: 0 (0%)
Realigned with forced non-floating right bound in left read: 0 (0%)
Realigned with forced non-floating left bound in right read: 0 (0%)
====================

[mizraelson]

Hi,

From the report, we observe that only 18% of the reads cover the CDR3 region. Additionally, about 25% of reads are missing the J gene, and a significant 57% of reads are non-target.

In MiXCR, if a read solely covers the C gene without even partially covering the J gene, it's categorized as off-target. Reads fitting this description, if present, should appear in the Infected_unaligned.fastq. Unfortunately, as of now, it's not feasible to align reads that only encompass the C gene. We're looking to amend this in future updates to provide more flexibility in this aspect.

Regarding the 25% of reads lacking the J gene: By integrating the --keep-non-CDR3-alignments parameter into your command, these reads will be retained in the '.vdjca' file. This will allow you to manually inspect them using the mixcr exportAlignmentsPretty command. Here's an example:

mixcr exportAlignmentsPretty \
    -s 100 \ 
    -n 20 \
    Mixcr_demul_Infected.vdjca

Executing the above command will skip the initial 100 alignments and subsequently export the next 20. This should suffice for identifying the presence of such reads.

[ShaowenJ]

Hi @mizraelson
Thanks for the clear explanation again.
I will try on the --keep-non-CDR3-alignments parameter.
Another quick question if you don't mind, I noticed you have a new preset for Nanopore long-read sequencing at version 4.2.2, is there any big difference between the new one and the old one?
And how do you guys decide about the scoring&penalty parameters? I am just wondering if I should play around the parameter to get more output.

Thanks again!

[ShaowenJ]

Very Interesting, with the --keep-non-CDR3-alignments parameter
Most of the reads aligned to TRAD.

Analysis time: 31.89s
Total sequencing reads: 127641
Successfully aligned reads: 63623 (49.85%)
Alignment failed, no hits (not TCR/IG?): 64018 (50.15%)
Overlapped: 0 (0%)
Overlapped and aligned: 0 (0%)
Alignment-aided overlaps: 0 (NaN%)
Overlapped and not aligned: 0 (0%)
No CDR3 parts alignments, percent of successfully aligned: 36404 (57.22%)
Partial aligned reads, percent of successfully aligned: 4893 (7.69%)
TRA chains: 434 (0.68%)
TRA non-functional: 42 (9.68%)
TRB chains: 3050 (4.79%)
TRB non-functional: 65 (2.13%)
TRD chains: 9 (0.01%)
TRD non-functional: 2 (22.22%)
TRG chains: 103 (0.16%)
TRG non-functional: 2 (1.94%)
TRAD chains: 18607 (29.25%)
TRAD non-functional: 0 (0%)
IGH chains: 9160 (14.4%)
IGH non-functional: 392 (4.28%)
IGK chains: 26315 (41.36%)
IGK non-functional: 533 (2.03%)
IGL chains: 5945 (9.34%)
IGL non-functional: 999 (16.8%)
Realigned with forced non-floating bound: 0 (0%)
Realigned with forced non-floating right bound in left read: 0 (0%)
Realigned with forced non-floating left bound in right read: 0 (0%)

What does TRAD mean here? Have you guys encountered this before?

[mizraelson]

The preset you're using is the same for the most part. However, it might be beneficial to switch to the latest version, as we've implemented several algorithmic improvements since then.

Regarding the scoring, we've extensively tested it on a wide range of datasets, so it's best to leave it as is. Typically, that's not the root of the issue.

The TRAD label is used when we can't confidently determine if it's TRA or TRD, because these chains share some V genes. This aligns with your observation about some alignments missing the J gene; apparently many of these are associated with TRAD V gene alignments. I'd suggest reviewing the alignments to identify the sequence that follows the V gene.
Given that many of the reads didn't align, it would be worthwhile to inspect them manually with BLAST. If you find non-aligned reads that come from the TCR/BCR receptor sequence, please let me know, and we'll investigate further.

If you're open to sharing the raw file, or a portion of it, I can give it a closer look for you.

[ShaowenJ]

Hi @mizraelson
Thanks so much for helping us to dig into this. The observations are very helpful and interesting! We will try to get the newest version of MiXCR (v4.3.2) to realign the sequence.
About the unaligned reads, could I ask how you do the BLAST, I am not too familiar with BLAST in a high-throughput way. Could you give me some general direction? That would be very grateful.
If you don't want to share here, you can send it to my email: sj657@cornell.edu

[mizraelson]

Hi,

The latest version of MiXCR, 4.4.2, is available for download here.

I typically utilize this link to BLAST some non-target sequences. I simply paste the sequence, retain the default parameters, and click the BLAST button.

If a read is only partially aligned, I attempt to blast the segment that didn't align with the TCR/BCR reference.

Consider the command I run:

mixcr analyze generic-ont \
    --species mmu \
    --keep-non-CDR3-alignments \
    Infected_unaligned.fastq \
    out

This command retains the non-CDR3 alignments, allowing for manual inspection. Since you've observed a surge in TRA chains, I'll use a TCR alpha alignment for this illustration:

mixcr exportAlignmentsPretty \
--chains TRA \
-s 100 \
-n10 \
out.vdjca

Subsequently, I identify alignments that match solely with the J and C genes, but where the V gene and CDR3 region are absent.

>>> Read ids: 49394



           F  F  F  F  F  F  F  F  F  F  F  L  Q  Y  T  F  I  T  S  K  S  P  H  T  H  P  R
Quality   99999999999999999999999999999999999999999999999999999999999999999999999999999999
Target0 0 TTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTGCAATATACTTTTATTACTTCAAAGAGCCCACATACCCATCCACG 79  Score


             V  A  R  C  C  L  E  L  P  S  L  L  Y  I  V  *  I  E  S  H  S  V  Q  S  W  V
Quality    99999999999999999999999999999999999999999999999999999999999999999999999999999999
Target0 80 AGTGGCACGGTGTTGCCTGGAATTGCCTAGTCTGCTGTACATAGTATGAATTGAATCTCATAGTGTGCAGTCTTGGGTCT 159  Score


            *  L  Y  *  I  L  S  F  V  L  R  Q  G  F  S  T  S  L  S  L  A  C  N  L  L  Y  R
Quality     99999999999999999999999999999999999999999999999999999999999999999999999999999999
Target0 160 GACTTTATTAAATTTTGAGTTTTGTTTTGAGACAGGGTTTCTCTACATCACTTTCTCTGGCCTGTAACTTGCTATATAGA 239  Score


             S  P  G  S  *  I  L  C  C  F  C  H  L  S  D  K  N  Y  S  H  L  P  P  Q  V  T  *
Quality     99999999999999999999999999999999999999999999999999999999999999999999999999999999
Target0 240 TCACCGGGATCTTGAATTCTGTGTTGTTTCTGCCACTTGAGTGACAAAAATTACAGTCATTTGCCACCACAAGTGACCTA 319  Score


              F  T  I  A  R  S  V  H  L  C  A  L  F  W  G  E  D  L  E  M  Q  P  V  L  A  *
Quality     99999999999999999999999999999999999999999999999999999999999999999999999999999999
Target0 320 ATTTACCATTGCAAGGTCTGTGCATTTGTGTGCTCTGTTTTGGGGTGAGGATCTTGAAATGCAGCCAGTGCTGGCATAGA 399  Score


            S  C  *  P  C  W  C  R  K  R  V  W  H  A  D  T  S  V  L  C  P  Y  R  T  W  E  I
Quality     99999999999999999999999999999999999999999999999999999999999999999999999999999999
Target0 400 GCTGTTGACCGTGCTGGTGTAGAAAGAGAGTATGGCATGCCGACACCTCTGTTCTCTGTCCCTACCGTACCTGGGAGATA 479  Score

                                                   <J                   CDR3><FR4
               T  T  R  S  F  C  *  S  V  Y  Y  C  G  N  N  N  N  A  P  R  F  G  A  G  T  K  L
  Quality     99999999999999999999999999999999999999999999999999999999999999999999999999999999
  Target0 480 ACAACAAGAAGTTTTTGTTAGAGTGTGTATTACTGTGGCAATAACAACAATGCCCCACGATTTGGAGCGGGAACCAAATT 559  Score
TRAJ43*00  20                                      gcaataacaacaatgccccacgatttggagcgggaaccaaatt 62   570

                        FR4><C
                S  V  K  P  N  I  Q  N  P  E  P  A  V  Y  Q  L  K  D  P  R  S  Q  D  S  T  L
  Quality     99999999999999999999999999999999999999999999999999999999999999999999999999999999
  Target0 560 ATCAGTAAAACCAAACATCCAGAACCCAGAACCTGCTGTGTACCAGTTAAAAGATCCTCGGTCTCAGGACAGCACCCTCC 639  Score
TRAJ43*00  63 atcagtaaaaccaa                                                                   76   570
  TRAC*00   0               acatccagaacccagaacctgctgtgtaccagttaaaagatcctcggtctcaggacagcaccctcT 65   1291


            R  L  F  T  D  F  D  S  Q  I  N  V  P  K  T  M  E  S  G  T  F  I  T  D  K  T  V
Quality     99999999999999999999999999999999999999999999999999999999999999999999999999999999
Target0 640 GCCTGTTCACCGACTTTGACTCCCAAATCAATGTGCCGAAAACCATGGAATCTGGAACGTTCATCACTGACAAAACTGTG 719  Score
TRAC*00  66 gcctgttcaccgactttgactcccaaatcaatgtgccgaaaaccatggaatctggaacgttcatcactgacaaaactgtg 145  1291


             L  D  M  K  A  M  D  S  K  S  N  G  A  I  A  W  S  N  Q  T  S  F  T  C  Q  D  I
Quality     99999999999999999999999999999999999999999999999999999999999999999999999999999999
Target0 720 CTGGACATGAAAGCTATGGATTCCAAGAGCAATGGGGCCATTGCCTGGAGCAACCAGACAAGCTTCACCTGCCAAGATAT 799  Score
TRAC*00 146 ctggacatgaaagctatggattccaagagcaatggggccattgcctggagcaaccagacaagcttcacctgccaagatat 225  1291


              F  K  E  T  N  A  T  Y  P  S  S _
Quality     99999999999999999999999999999999999999999999999999999999999999999999999999999999
Target0 800 CTTCAAAGAGACCAACGCCACCTACCCCAGTTCAGACGTTCCCTGTGATGCCACGTTGACTGAGAAAAGCTTTGAAACAG 879  Score
TRAC*00 226 cttcaaagagaccaacgccacctaccccagttcag                                              260  1291



Quality     99999999999999999999999999999999999999999999999999999999999999999999999999999999
Target0 880 ATATGAACCTAAACTTTCAAAACCTGTCAGTTATGGGACTCCGAATCCTCCTGCTGAAAGTAGCCGGATTTAACCTGCTC 959  Score



Quality     99999999999999999999999999999999999999999999999999999999999999999999999999999999
Target0 960 ATGACGCTGAGGCTGTGGTCCAGTTGAGGTCTGCAAGACTGACAGAGCCTGACTCCCAAGCTCCATCCTCCTCACCCCTC 1039  Score

When a sequence doesn't align to any known V genes, it's important to explore its potential origins to understand the context of the data better.

Given the provided unaligned part of the sequence:

TTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTGCAATATACTTTTATTACTTCAAAGAGCCCACATACCCATCCACGAGTGGCACGGTGTTGCCTGGAATTGCCTAGTCTGCTGTACATAGTATGAATTGAATCTCATAGTGTGCAGTCTTGGGTCTGACTTTATTAAATTTTGAGTTTTGTTTTGAGACAGGGTTTCTCTACATCACTTTCTCTGGCCTGTAACTTGCTATATAGATCACCGGGATCTTGAATTCTGTGTTGTTTCTGCCACTTGAGTGACAAAAATTACAGTCATTTGCCACCACAAGTGACCTAATTTACCATTGCAAGGTCTGTGCATTTGTGTGCTCTGTTTTGGGGTGAGGATCTTGAAATGCAGCCAGTGCTGGCATAGAGCTGTTGACCGTGCTGGTGTAGAAAGAGAGTATGGCATGCCGACACCTCTGTTCTCTGTCCCTACCGTACCTGGGAGATAACAACAAGAAGTTTTTGTTAGAGTGTGTATTACTGTG

When I BLASTed it, I figured it aligns well with the mouse gene Gtf3c6. Gtf3c6 is not a TCR or BCR gene, suggesting the sequence might be from a non-specific amplification or other processes during the sequencing workflow. It could be from off-target capture or other artifacts.

I see it's a perfect match, except for the last 93 nucleotides which are missing:

AGAGAGTATGGCATGCCGACACCTCTGTTCTCTGTCCCTACCGTACCTGGGAGATAACAACAAGAAGTTTTTGTTAGAGTGTGTATTACTGTG

Then I BLAST this part separately:

It appears to align within the TRA locus. When I examine the Graphics of one of the alignments, I notice that this sequence actually originates from the DNA downstream of the TRAJ43 gene (the DNA that is typically excised during VDJ recombination):

These are the fundamental steps to delve into the issue. Please let me know if you have any questions.

[ShaowenJ]

Thanks so much! You are awesome

[ShaowenJ]

Hi @mizraelson , sorry to bring this up, I just have a quick follow-up question regarding the alignment results
So the reads were assigned as "Alignment failed: no hits (not TCR/IG?)": It could mean that the read can be partially aligned to VDJ-C segments, but not the alignment for whole complete segments or no CDR3 region. Because I did some BLAST for the unaligned reads, it seems that only 20% from the unaligned reads cannot be aligned to the constant region of BCR/TCR

Thanks for the help again!

[mizraelson]

It depends on the preset and parameters used. For some presets, the message 'Alignment failed: no hits (not TCR/IG?)' indeed indicates that these reads did not completely cover the CDR3 region. This behavior is expected in protocols where a read or a pair of reads is anticipated to fully cover CDR3 region, as is the case with the Nanopore preset. However, other presets, such as RNA-seq, may include the --keep-non-CDR3-alignments option by default. In such cases, these reads would be counted among the successfully aligned ones, even if they only partially cover CDR3 or cover only V or J genes. Currently, reads that only cover C gene are considered as Alignment failed: no hits (not TCR/IG?) for any preset.