Cholangiocarcinoma CHOL
Of 60,498 genes, 4,915 are higher in the tumor than in normal liver and in whole blood, and are not made by blood immune cells. 2,141 are protein-coding, and 2,130 of those have supporting plasma Cell-free RNA (cfRNA)RNA fragments that circulate in blood plasma outside cells. Most come from blood cells; a small share comes from other tissues, including tumors. evidence; the rest are mostly non-coding genes the plasma sources cannot check.
Samples compared
- Tumor
- 36
- TCGA primary tumors
- Normal tissue
- 119
- 9 TCGA tumor-adjacent, 110 GTEx liver
- Whole blood
- 337
- GTEx samples from 328 donors
- Survival
- 36
- patients, 20 events (Survival endpointsCurated TCGA outcomes (Liu et al. 2018): OS is overall survival, PFI the progression-free interval and DSS disease-specific survival. LB2 uses whichever has the most events in each cancer type.)
Results in brief
- 4,915
- candidate genes: 2,130 with plasma cfRNA evidence, 32 protein-coding without, 2,753 non-coding
- 64
- genes in the tissue panel, Held-out AUCHow well an elastic-net model separates tumor from normal tissue, measured with nested cross-validation on samples the model did not train on. It describes tissue, not the accuracy of a blood test. 0.999
- 0
- genes associated with progression-free interval (FDR < 0.05, 150 tested)
Candidates
Filter by plasma evidence or discrimination, find a gene, and select any point or row for its full record. The table holds all 4,915 candidates.
How the candidates were selected
60,498 genes tested; 17,768 higher in tumor than in normal liver; 10,852 also higher than in whole blood; 4,915 remain after removing genes expressed in blood immune cells; 2,130 of these have plasma cfRNA evidence, 32 protein-coding candidates have none, and 2,753 are non-coding genes the plasma sources cannot list.
Each comparison uses a two-sided Wilcoxon rank-sum test. A gene passes when its q-value is below 0.05, its log2 fold change is at least 1, and the lower bound of the 95% confidence interval for its AUC is at least 0.70.
Genes made by any of 18 sorted blood immune cell types above 1 nTPM are then removed, because blood cells supply most plasma RNA. Genes missing from that reference, mostly non-coding, are kept.
Plasma evidence only reorders the list. A candidate not yet seen in plasma stays in, because every dataset misses genes. Read the full methods.
Tissue classifier panel
- FOXM1+0.12
- RP11-2C24.9+0.09
- BAIAP2L2+0.09
- NFE2L3+0.09
- PPP1R37+0.08
- VN1R1+0.08
- ANKRD52+0.07
- B4GALNT2+0.07
- TFAP4+0.07
- LRRC56+0.07
- MRGBP+0.07
- MYEF2+0.07
- ZNF48+0.06
- ABHD17C+0.06
- C19orf48+0.06
- COLCA2+0.06
- ELMO3+0.06
- RHPN1+0.05
- ATAT1+0.05
- CMTM4+0.05
Show the other 44 genes
- LAMA5+0.05
- ITPR3+0.05
- TIGD5+0.05
- B4GALNT4+0.04
- NAT14+0.04
- PPP1R14C+0.04
- STIL+0.04
- FIZ1+0.04
- CLIP2+0.04
- SETD1A+0.03
- SOCS7+0.03
- LAD1+0.03
- PIP5K1C+0.03
- LAMB3+0.02
- EHMT2+0.02
- THUMPD2+0.02
- SCX+0.02
- ZNF286B+0.02
- DUOX1+0.02
- AGAP5+0.02
- CDC42BPG+0.02
- MAPK11+0.02
- ZNF653+0.02
- RAB40A+0.01
- C6orf48+0.01
- SENP3+0.01
- RAI1+0.01
- DLG3+0.01
- CRMP1+0.01
- ZNF579+0.01
- ANKS6+0.01
- ABCA3+0.00
- APLP1+0.00
- PRR36+0.00
- PRDM11+0.00
- CCNE1+0.00
- RGS17+0.00
- PLCD3+0.00
- HSD11B2+0.00
- GULP1+0.00
- SEPT3+0.00
- ARHGAP8+0.00
- INPP5J+0.00
- WIZ+0.00
An elastic-net logistic regression was trained on the 300 top-ranked candidates to tell tumor from normal tissue. Its Held-out AUCHow well an elastic-net model separates tumor from normal tissue, measured with nested cross-validation on samples the model did not train on. It describes tissue, not the accuracy of a blood test. is 0.999, from nested five-fold cross-validation.
Before training, the TCGA-versus-GTEx offset was removed from the expression matrix with the tumor/normal contrast protected, as in the tissue comparison. That correction is fitted once on all samples, so the held-out AUC is not fully independent of it.
This number describes tumor and normal tissue, where separation is expected to be near perfect. It is not the accuracy of a blood test: that has to be measured in plasma from patients and controls.
Association with survival
None of the 150 top-ranked candidates is associated with progression-free interval at FDR < 0.05 (36 patients, 20 events). A null result here is informative: these genes mark the presence of the cancer, not necessarily its course.
For the 150 top-ranked candidates, a Cox model relates tumor expression, as a continuous value, to progression-free interval in 36 patients (20 events). No high/low cutpoint is searched for, since optimized cutpoints inflate false positives.
A Hazard ratio per SDFrom a Cox model with the gene’s tumor expression as a continuous variable: the change in hazard for each one-standard-deviation increase. Above 1, higher expression goes with a shorter time to the event; below 1, with a longer one. above 1 means higher expression goes with a shorter progression-free interval. These are associations in tissue, not evidence that a blood level predicts outcome.
Caveats for CHOL
- GTEx has no bile duct, so normal liver is used alongside the TCGA adjacent samples.
- Only 9 TCGA adjacent samples anchor the correction for study differences between TCGA and GTEx normals.
- With only 36 tumors, confidence intervals are wide; treat close ranks as ties.
- Only 20 PFI events, so survival estimates are imprecise.
- The blood comparison sets TCGA tumors against GTEx blood. Study and biology cannot be separated there, which is why genes made by blood immune cells are removed as well.