TRUST4 assembles T-cell and B-cell receptor sequences — including the hypervariable CDR3 — directly from the RNA-seq reads, reconstructing the immune cells that were present in each sample. It was run on the same three RNA BAMs as the rest of this site (the hand primary, the lung metastasis, and a normal-lung baseline), and independently reproduces the picture from oncoanalyser's CIDER: the hand primary is nearly devoid of B cells while both lung samples carry large, class-switched, B-cell–dominated repertoires.
The question this page explores: among the antibody (IG) clones in the lung metastasis, are any enriched there relative to normal lung — a possible, though unproven, signature of a local response to the tumor? The short answer is that the two lung repertoires overlap only slightly (Jaccard 0.045), but most of that is shallow sampling of a very diverse repertoire, not biology. The useful output is therefore not the raw "met-only" list but a ranked shortlist of expanded, class-switched, met-enriched clones — leads to follow up, not conclusions.
Productive CDR3 clonotypes assembled per locus. The hand primary is nearly devoid of B-cell (IG) clones and carries only a few T-cell (TR) clonotypes; both lung samples carry large, B-cell–dominated repertoires (IGK > IGL ≈ IGH ≫ T cells). Counts are normalized to library size (per million Isofox fragments) so samples of different depth compare fairly.
Heavy-chain isotype of each B-cell clone. Both lung repertoires are dominated by IgA (the mucosal default in lung) with a class-switched IgG component and little naive IgM — an antigen-experienced, class-switched profile.
A tumor-driven B-cell response would show as more clonal expansion or more somatic hypermutation at the metastasis. Here the two lung repertoires are similar — the metastasis is not more oligoclonal or more mutated than normal lung — an important caveat for interpreting the clone overlap below.
IGH clones (by CDR3 amino acid) shared between the two same-batch lung samples vs private to each. Overlap is low (Jaccard 0.045) — expected when a diverse repertoire is sampled shallowly, so most "met-only" clones are sampling, not biology. The interesting subset is the ranked candidates below: expanded, class-switched, met-enriched clones.
Ranked by a transparent composite of expansion × class-switch × met-enrichment (shown as Score). These are hypothesis-generating leads, not validated tumor-reactive antibodies — sequence alone cannot prove specificity.
| CDR3 (aa)ⓘ | Vⓘ | Isotypeⓘ | Met readsⓘ | Normalⓘ | SHM %simⓘ | Signalsⓘ | Scoreⓘ ▾ |
|---|---|---|---|---|---|---|---|
| CARDSSGYGRGFGGRYYYGMDVW | IGHV3-33·IGHJ6 | IGHA1 | 213 | 0 | 95.6 | met-only | 554 |
| CARGTGSHRSGWPYFDYW | IGHV3-21·IGHJ4 | IGHA1 | 140 | 0 | 97.1 | met-only | 364 |
| CARERRRDVWGSYRYRKAFDIW | IGHV3-23·IGHJ3 | IGHA1 | 87 | 22 | 93.8 | convergent | 339 |
| CARISPGGHGFDVW | IGHV1-18·IGHJ3 | IGHG3 | 71 | 0 | 93.1 | met-only IgG convergent | 256 |
| CAKVLSLTENYWYGMDVW | IGHV3-23·IGHJ6 | IGHA1 | 49 | 13 | 94.7 | 191 | |
| CTRDWGVVVLPAAMGFDYW | IGHV3-49·IGHJ4 | IGHA1 | 49 | 6 | 97.8 | convergent | 191 |
| CAKDWGGLLMNNWYFDLW | IGHV3-23·IGHJ2 | IGHA1 | 45 | 12 | 97.5 | 176 | |
| CARGDWFDPW | IGHV1-3·IGHJ5 | IGHA1 | 44 | 4 | 97.1 | convergent | 172 |
| CVREGYGDYGDYW | IGHV3-21·IGHJ4 | IGHA1 | 42 | 9 | 94.1 | convergent | 164 |
| CARQKIAAAGRRYYYYGMDVW | IGHV4-59·IGHJ6 | IGHA1 | 57 | 0 | 100.0 | met-only | 148 |
| CARKRRGDRYPFDFW | IGHV1-3·IGHJ4 | IGHG2 | 32 | 0 | 93.3 | met-only IgG | 115 |
| CARNAYFYGWYLDLW | IGHV3-48·IGHJ2 | IGHG3 | 21 | 6 | 90.0 | IgG | 113 |
| CARDRVHGMDVW | IGHV1-18·IGHJ6 | IGHM | 35 | 2 | 94.7 | 105 | |
| CARDRYYDFWSGYFDYW | IGHV3-11·IGHJ4 | IGHA1 | 38 | 0 | 100.0 | met-only | 99 |
| CASGPMTTVTTDWYFDLW | IGHV5-51·IGHJ2 | IGHA1 | 25 | 4 | 97.4 | 98 | |
| CTTPSPSYQLRRDYW | IGHV3-15·IGHJ4 | IGHA1 | 35 | 0 | 96.2 | met-only | 91 |
| CARLSVPTAFGVISHYDFW | IGHV1-2·IGHJ4 | IGHG3 | 25 | 0 | 100.0 | met-only IgG | 90 |
| CARDLGRWLQDAFDIW | IGHV1-2·IGHJ3 | IGHA1 | 24 | 9 | 100.0 | 74 | |
| CAKDRAWVRSPRGDYFDYW | IGHV3-23·IGHJ4 | IGHA1 | 19 | 3 | 93.8 | convergent | 74 |
| CARDIFPHSSSFLYYFDYW | IGHV3-21·IGHJ4 | IGHA1 | 28 | 0 | 94.7 | met-only | 73 |
| CARGIYGDFPYFDYW | IGHV4-38-2·IGHJ4 | IGHA1 | 27 | 0 | 97.1 | met-only | 70 |
| CARRAYCDGDCTLFNYLYFGLDVW | IGHV5-51·IGHJ6 | IGHG1 | 19 | 0 | 91.7 | met-only IgG | 68 |
| CTKDVEPGGATRW | IGHV3-9·IGHJ4 | IGHA2 | 26 | 0 | 88.9 | met-only | 68 |
| CVRDWQWLEGRDCFDPW | IGHV1-18·IGHJ5 | IGHA1 | 17 | 2 | 94.6 | 66 | |
| CARDLGKMAGASLDSW | IGHV3-48·IGHJ4 | IGHG1 | 17 | 0 | 92.3 | met-only IgG | 61 |
| CAKSDPNGVAAVFDYW | IGHV3-9·IGHJ4 | IGHG1 | 16 | 0 | 100.0 | met-only IgG | 58 |
| CASEAGYYYDSSGYYGDYW | IGHV3-21·IGHJ4 | IGHA2 | 22 | 0 | 95.2 | met-only | 57 |
| CAHRRETGDLFPPSPGFDYW | IGHV2-5·IGHJ4 | IGHA1 | 22 | 0 | 100.0 | met-only | 57 |
| CARIGLQHPYYYGMDVW | IGHV4-39·IGHJ6 | IGHA1 | 17 | 6 | 97.5 | 56 | |
| CARDNDGPVVDGYNGIGYW | IGHV3-11·IGHJ4 | IGHA1 | 20 | 0 | 94.1 | met-only | 52 |
Isofox gene expression (AdjTPM, log scale) for B-cell, plasma-cell and tertiary-lymphoid-structure (TLS) markers. This says whether the tissue even supports a local B-cell response. The metastasis is plasma-cell / IgA-secretion skewed (high JCHAIN) but has fewer T cells and a weaker TLS signature than normal lung.
Bars are log₁₀(TPM+1)-scaled; values are AdjTPM.
Please contact us with any questions or comments.