| Gene | Log2 Fold Change | P-Value | Adjusted P-Value (FDR) | Direction | Targetable Therapy |
|---|---|---|---|---|---|
| MYC | 3.0031 | 2.1580e-27 | 2.3307e-25 | UP | Yes (JQ1 (Clinical Trial), Molibresib (Clinical Trial)) |
| EGFR | 3.2383 | 8.4138e-24 | 4.5435e-22 | UP | Yes (Osimertinib, Erlotinib, Gefitinib, Cetuximab) |
| ESR1 | 4.1157 | 2.1583e-22 | 7.5055e-21 | UP | Yes (Tamoxifen, Fulvestrant, Elacestrant) |
| TP53 | -3.6491 | 2.7798e-22 | 7.5055e-21 | DOWN | Yes (PC14586 (Clinical Trial), Kevetrin (Clinical Trial)) |
| GENE_90 | -0.1749 | 9.6189e-06 | 2.0777e-04 | NS | No known drugs |
| GENE_29 | -0.1883 | 2.5261e-04 | 4.5469e-03 | NS | No known drugs |
| GENE_72 | -0.1669 | 3.2001e-04 | 4.9373e-03 | NS | No known drugs |
| GENE_17 | -0.1481 | 3.6993e-04 | 4.9941e-03 | NS | No known drugs |
| GENE_28 | -0.1156 | 1.2968e-03 | 1.4006e-02 | NS | No known drugs |
| GENE_45 | -0.1101 | 1.1796e-03 | 1.4006e-02 | NS | No known drugs |
| GENE_20 | -0.1467 | 2.5766e-03 | 2.3189e-02 | NS | No known drugs |
| GENE_73 | -0.1282 | 2.4575e-03 | 2.3189e-02 | NS | No known drugs |
| GENE_95 | -0.1349 | 3.0798e-03 | 2.5586e-02 | NS | No known drugs |
| GENE_8 | -0.1723 | 3.3307e-03 | 2.5694e-02 | NS | No known drugs |
| GENE_65 | -0.1271 | 4.0986e-03 | 2.9219e-02 | NS | No known drugs |
| GENE_70 | -0.1220 | 4.3287e-03 | 2.9219e-02 | NS | No known drugs |
| GENE_80 | -0.1251 | 7.7265e-03 | 4.9086e-02 | NS | No known drugs |
| GENE_43 | -0.1262 | 1.1342e-02 | 6.4471e-02 | NS | No known drugs |
| GENE_10 | -0.1178 | 1.1119e-02 | 6.4471e-02 | NS | No known drugs |
| GENE_33 | -0.1179 | 1.2157e-02 | 6.5650e-02 | NS | No known drugs |
| GENE_71 | -0.1208 | 1.3080e-02 | 6.7268e-02 | NS | No known drugs |
| GENE_81 | -0.1010 | 1.3932e-02 | 6.8392e-02 | NS | No known drugs |
| GENE_63 | -0.1254 | 1.6719e-02 | 7.8507e-02 | NS | No known drugs |
| GENE_87 | -0.1119 | 1.7743e-02 | 7.9842e-02 | NS | No known drugs |
| GENE_49 | -0.1313 | 1.8845e-02 | 8.1121e-02 | NS | No known drugs |
| GENE_30 | -0.0972 | 1.9529e-02 | 8.1121e-02 | NS | No known drugs |
| GENE_12 | -0.1236 | 2.2086e-02 | 8.8344e-02 | NS | No known drugs |
| GENE_18 | -0.1066 | 2.3044e-02 | 8.8882e-02 | NS | No known drugs |
| GENE_19 | -0.0932 | 2.4699e-02 | 9.1984e-02 | NS | No known drugs |
| GENE_51 | -0.1010 | 2.5840e-02 | 9.3023e-02 | NS | No known drugs |
| GENE_5 | -0.1015 | 2.9506e-02 | 1.0279e-01 | NS | No known drugs |
| GENE_42 | -0.1007 | 3.2517e-02 | 1.0329e-01 | NS | No known drugs |
| GENE_99 | -0.0973 | 3.1695e-02 | 1.0329e-01 | NS | No known drugs |
| GENE_83 | -0.0968 | 3.1179e-02 | 1.0329e-01 | NS | No known drugs |
| GENE_39 | -0.1091 | 3.4550e-02 | 1.0661e-01 | NS | No known drugs |
| GENE_94 | -0.0993 | 3.6643e-02 | 1.0696e-01 | NS | No known drugs |
| GENE_36 | -0.0905 | 3.5766e-02 | 1.0696e-01 | NS | No known drugs |
| GENE_25 | -0.1025 | 3.9178e-02 | 1.1135e-01 | NS | No known drugs |
| GENE_1 | -0.1030 | 4.2289e-02 | 1.1711e-01 | NS | No known drugs |
| GENE_75 | -0.1024 | 4.4190e-02 | 1.1830e-01 | NS | No known drugs |
| GENE_54 | -0.0781 | 4.4912e-02 | 1.1830e-01 | NS | No known drugs |
| GENE_97 | -0.0901 | 4.6872e-02 | 1.2053e-01 | NS | No known drugs |
| GENE_55 | -0.0964 | 4.8958e-02 | 1.2296e-01 | NS | No known drugs |
| CDK4 | -0.1017 | 5.3098e-02 | 1.3033e-01 | NS | Yes (Palbociclib, Ribociclib, Abemaciclib) |
| BRCA1 | -0.0805 | 5.6629e-02 | 1.3591e-01 | NS | Yes (Olaparib, Niraparib, Rucaparib, Talazoparib) |
| GENE_77 | -0.0865 | 5.8618e-02 | 1.3763e-01 | NS | No known drugs |
| GENE_23 | -0.1012 | 6.0827e-02 | 1.3977e-01 | NS | No known drugs |
| GENE_48 | -0.0997 | 6.7077e-02 | 1.5050e-01 | NS | No known drugs |
| GENE_57 | -0.0823 | 6.9502e-02 | 1.5050e-01 | NS | No known drugs |
| GENE_91 | -0.0731 | 6.9674e-02 | 1.5050e-01 | NS | No known drugs |
RNA-seq count matrices are queried from the DuckDB + Parquet local database. Counts are normalized for sequencing library depth (CPM) or processed using PyDESeq2's negative binomial Generalized Linear Model. If PyDESeq2 convergence fails or sample size is small, a fallback Welch's t-test on log-transformed counts is performed. P-values are adjusted for multiple testing using the Benjamini-Hochberg (FDR) procedure.
If multiple sequencing plates or batches are identified in clinical metadata, an empirical location-and-scale batch correction (ComBat) is applied to prevent batch artifacts from skewing results.
For pan-cancer gene validation, the CCCS combines Directionality (w=0.25), Fold Change Magnitude (w=0.25), Survival Significance (w=0.35), and Statistical Significance (w=0.15). Scores are scaled by the consistency ratio of significant cancer types.