KRAS p.Gly12Cys

Variant summary

Our verdict is Pathogenic.
>+14 Pathogenic
-7
-6
-1
0
+5
+6
+9
+10
B
LB
VUS
LP
P
The variant received 27 classification points (ACGS-UK Somatic Oncogenicity v2025). O1_Very_StrongO3_ModerateO4_ModerateO5_SupportingO6_SupportingO7_StrongO8_Supporting

The NM_004985.5(KRAS):c.34G>T (p.Gly12Cys) variant causes a missense change involving the alteration of a conserved nucleotide. The gene KRAS is a known oncogene (CancerMine: 2 TSG, 2158 oncogene, 346 driver citations). The gene KRAS is a cancer driver gene (CancerMine: 2 TSG, 2158 oncogene, 346 driver citations). The variant is absent from the gnomAD population database at sites with sufficient sequencing coverage. In-silico predictor (REVEL) classifies this variant as likely damaging/oncogenic. Splicing prediction tools (SpliceAI) predict no significant impact on normal splicing. The affected nucleotide is highly conserved across species (PhyloP 100-way vertebrate score: 7.89). Variant has been reported in ClinVar as Pathogenic/Likely Pathogenic (★). ClinVar reports functional evidence for this variant: "SCV004501632: Experimental studies have shown that this missense change affects KRAS function (PMID:16051643).". A different missense at the same amino acid position has been reported as Pathogenic/Likely Pathogenic in ClinVar: p.G12R: Pathogenic/Likely_pathogenic (ClinVar VariationId 12579, 2 stars) Other variants at the same amino acid position have been reported in ClinVar (not pathogenic): p.G12= (synonymous): Likely_benign (ClinVar VariationId 1734094, 1 star); p.G12= (synonymous): Likely_benign (ClinVar VariationId 4859020, 1 star); p.G12= (synonymous): Likely_benign (ClinVar VariationId 1734090, 1 star) This variant is listed in the SVIG-UK Canonical Variants List — a curated registry of well-established oncogenic variants (O1 criterion, stand-alone Oncogenic). This exact variant is established as oncogenic in: ClinVar, CGI (Cancer Genome Interpreter). The variant position is a cancer hotspot (cancerhotspots.org): 2910 samples carry this exact amino acid change out of 32010 samples with a variant at this residue. The variant has been observed in cBioPortal in 5927 samples across 145 studies and 134 cancer types; somatic enrichment level: MODERATE (Observed in many cases across multiple studies/cancer contexts.). This exact variant is curated in the UniProt human variants database as Uncertain Significance, associated with Juvenile myelomonocytic leukemia (jmml); it is also listed as a COSMIC curated somatic variant.

Frequency

Genomes: not found (cov: 32)

Consequence

KRAS
NM_004985.5 missense

Scores

12
6
1

Clinical Significance

Likely pathogenic criteria provided, single submitter P:7O:6

Conservation

PhyloP100: 7.89

Publications

1667 publications found
Variant links:
Genes affected
KRAS (HGNC:6407): (KRas proto-oncogene, GTPase) This gene, a Kirsten ras oncogene homolog from the mammalian ras gene family, encodes a protein that is a member of the small GTPase superfamily. A single amino acid substitution is responsible for an activating mutation. The transforming protein that results is implicated in various malignancies, including lung adenocarcinoma, mucinous adenoma, ductal carcinoma of the pancreas and colorectal carcinoma. Alternative splicing leads to variants encoding two isoforms that differ in the C-terminal region. [provided by RefSeq, Jul 2008]
KRAS Gene-Disease associations (from GenCC):
  • cardiofaciocutaneous syndrome 2
    Inheritance: AD Classification: DEFINITIVE, STRONG, MODERATE Submitted by: Genomics England PanelApp, Labcorp Genetics (formerly Invitae), PanelApp Australia, Ambry Genetics, G2P
  • Noonan syndrome
    Inheritance: AD Classification: DEFINITIVE, SUPPORTIVE Submitted by: Orphanet, ClinGen
  • Noonan syndrome 3
    Inheritance: AD Classification: DEFINITIVE, STRONG Submitted by: G2P, PanelApp Australia, Labcorp Genetics (formerly Invitae), Ambry Genetics, Genomics England PanelApp
  • autoimmune lymphoproliferative syndrome type 4
    Inheritance: AD Classification: STRONG Submitted by: PanelApp Australia
  • cardiofaciocutaneous syndrome
    Inheritance: AD Classification: STRONG, SUPPORTIVE Submitted by: Orphanet, ClinGen
  • linear nevus sebaceous syndrome
    Inheritance: AD Classification: STRONG Submitted by: Labcorp Genetics (formerly Invitae)
  • Costello syndrome
    Inheritance: AD Classification: NO_KNOWN Submitted by: ClinGen

Genome browser will be placed here

new If you want to explore the variant's impact on the transcript NM_004985.5, check out the Mutation Effect Viewer. This is especially useful for frameshift variants or if you want to visualize the effect of exon loss / intron retention.

Classification according to ACGS-UK Somatic Oncogenicity v2025

Classification was made for transcript:

Our verdict: Pathogenic. The variant received 27 points.

O1Established oncogenic variant
Variant on SVIG-UK Canonical Variants List — Stand-alone Oncogenic (O1); ClinVar oncogenic AND CGI oncogenic — two independent sources, Very Strong (O1); CGI exact DNA-change oncogenic — Strong (O1); ClinVar oncogenic/SCI T1 (single/no review, ★0–1) — Moderate (O1); O1 contributing sources (16.0 pts): SVIG-UK Canonical Variants List, ClinVar (somatic oncogenic/LO (★)), CGI (Cancer Genome Interpreter)
O3Absent or very rare in population
Absent or extremely rare in gnomAD (AC ≤ 2) — O3 moderate; GnomAD: AF = absent, AC = 0 — absent/extremely rare (O3 moderate [+2])
O4Enriched in somatic variant database
Enriched in cBioPortal (pre-computed score ≥ 2.0) — Moderate (O4); cBioPortal: samples=5927 | studies=145 | cancer types=134 | level=MODERATE | points=2.0 | reason: Observed in many cases across multiple studies/cancer contexts.
O5Same location/impact as known oncogenic
Different AA at same residue with oncogenic evidence (or germline P/LP in TSG), or same AA change SCI Tier II Potential — Supporting (O5); ClinVar same-AA oncogenic/T1: false | ClinVar same-AA SCI tier: — | CGI same-AA oncogenic: false | TSG: false | ClinVar same-AA germline P/LP: true | CIViC LOF: false | ClinVar diff-AA oncogenic: true | ClinVar diff-AA germline P/LP: true
O6Computational evidence supports deleterious effect
Computational evidence supports a deleterious/oncogenic effect; Missense variant — primary computational scorer supports an oncogenic/deleterious effect
O7Mutational hotspot or critical domain
Strong hotspot (cancerhotspots.org): >50 total samples, ≥10 same AA change; Note: exclude MSK/TCGA-derived samples from cancerhotspots count to avoid double-counting with O4; cancerhotspots.org: total samples at AA position = 32010, same AA change = 2910; Exclude MSK/TCGA-derived samples from count to avoid double-counting with O4; UniProt domain: 0 pathogenic, 0 benign.; ±8 AA neighbourhood: 16 pathogenic, 0 benign (OR vs. background: 165.0); Variant does not impact splicing — splice-hotspot path not applicable
O8Missense constraint in gene or region
Missense in gene/domain constrained for missense variation where missense is common disease mechanism (O8 supporting); Gene/disease mechanism: missense is a common mechanism of oncogenesis; gnomAD missense Z-score: 3.09 (threshold ≥3.09 for constraint); Domain-level constraint: variant is in a critically constrained functional domain (IS_AT_CRITICAL_DOMAIN)

Variant Effect in Transcripts

Automated classification analysis was done for transcript: NM_004985.5. You can select a different transcript below to see updated classification assignments.

RefSeq Transcripts

KRAS NM_033360.4
MANE Plus Clinical
HGVSc / HGVSpc.34G>Tp.Gly12Cys
Exon rankExon 2 of 6
missense
Prot: NP_203524.1UniProt: P01116-1
KRAS NM_004985.5
MANE Select
HGVSc / HGVSpc.34G>Tp.Gly12Cys
Exon rankExon 2 of 5
missense
Prot: NP_004976.2
KRAS NM_001369786.1
HGVSc / HGVSpc.34G>Tp.Gly12Cys
Exon rankExon 2 of 6
missense
Prot: NP_001356715.1UniProt: P01116-1

Ensembl Transcripts

KRAS ENST00000256078.10
TSL:1MANE Plus Clinical
HGVSc / HGVSpc.34G>Tp.Gly12Cys
Exon rankExon 2 of 6
missense
Prot: ENSP00000256078.5UniProt: P01116-1
KRAS ENST00000311936.8
TSL:1MANE Select
HGVSc / HGVSpc.34G>Tp.Gly12Cys
Exon rankExon 2 of 5
missense
Prot: ENSP00000308495.3UniProt: P01116-2
KRAS ENST00000556131.2
TSL:1
HGVSc / HGVSpc.34G>Tp.Gly12Cys
Exon rankExon 2 of 3
missense
Prot: ENSP00000451856.1UniProt: G3V4K2

Frequencies

Allele frequencies (AF), counts (AC/AN), homozygotes and coverage

Common (AF > 0.05 / Hom > 5)
Rare (AF ≤ 0.0001 / Hom ≤ 1)
Global population databases 5 sources
GnomAD3 genomes
Cov:32
GnomAD2 exomes
AF: 0.00
AC:0
AN:249272
GnomAD4 exome
Cov:31
GnomAD4 genome
Cov:32
ALFA (dbGaP/dbSNP Allele Frequency Aggregator)
AF: 0.00
AC:0
Hom:0
AN:55336
Showing 5 sources

ClinVar

ClinVar submissions
Significance:Likely pathogenic
Revision:criteria provided, single submitter
View on ClinVar
Pathogenic
VUS
Benign
Condition
1
-
-
Gallbladder cancer (1)
1
-
-
Lung adenocarcinoma (1)
1
-
-
Lung cancer (1)
1
-
-
Lung carcinoma (1)
1
-
-
Non-small cell lung carcinoma (1)
1
-
-
not provided (1)
1
-
-
RASopathy (1)
-
-
-
Adenocarcinoma of the large intestine (1)
-
-
-
Diffuse pediatric-type high-grade glioma, H3-wildtype and IDH-wildtype (1)
-
-
-
Embryonal rhabdomyosarcoma (1)
-
-
-
Endometrial carcinoma (1)
-
-
-
Neoplasm (1)
-
-
-
Ovarian mucinous adenocarcinoma (1)

Computational Scores

AlphaGenome AVI
Pathogenic32
AlphaMissense
Pathogenic1.0
BayesDel_addAF
Pathogenic0.39D
BayesDel_noAF
Pathogenic0.33
CADD
Pathogenic29
DANN
Uncertain1.0
DEOGEN2
Pathogenic0.87D
Eigen
Pathogenic0.95
Eigen_PC
Pathogenic0.91
FATHMM_MKL
Pathogenic0.98D
FuncVEP CTI
Pathogenic0.97
GPN-Star LLR
N/A-11
GPN-Star score
N/A11
LIST_S2
Benign0.55T
M_CAP
Uncertain0.15D
MetaRNN
Pathogenic0.91D
MetaSVM
Uncertain0.55D
Mutation Taster
N/A0/100disease causing (ClinVar)
MutationAssessor
Uncertain2.6M
PhyloP100
Pathogenic7.9
popEVE
Uncertain-4.8
PrimateAI
Pathogenic0.92D
PROVEAN
Pathogenic-7.0D
REVEL
Pathogenic0.85
Sift
Uncertain0.0060D
Sift4G
Uncertain0.022D
Varity_R
N/A0.94
VESM-3B
Benign-14
Showing 28 of 28 scores

Splicing Scores

Pangolin (max)
Benign0.010
SpliceAI score (max)
Benign0.0
Details are displayed if max score is > 0.2
Showing 2 of 2 scores

Find out detailed SpliceAI scores and Pangolin per-transcript scores at spliceailookup.broadinstitute.org

MaxEntScan Visualizer can be used to analyze the impact of this mutation on the neighboring sequence.

MaveDB functional evidence

MaveDB is a public repository of laboratory assays that test the effect of thousands of variants at once — treat this as supporting research data and interpret it with great caution; it is not a clinical classification.

14 no calibration
Protein level
binding fitness from KRAS-SOS1 bindingPCA of KRAS block1
The energetic and allosteric landscape for KRAS inhibition
Gene
KRAS
Score
-0.00609 · p.Gly12Cys (assay numbering)
No calibration data
This score is at approximately the 92th percentile of all variants tested in this assay.
This score has not been calibrated against ACMG evidence — shown for context only, not as clinical evidence.
Protein level
abundance fitness from abundancePCA of KRAS block1
The energetic and allosteric landscape for KRAS inhibition
Gene
KRAS
Score
-0.244 · p.Gly12Cys (assay numbering)
No calibration data
This score is at approximately the 68th percentile of all variants tested in this assay.
This score has not been calibrated against ACMG evidence — shown for context only, not as clinical evidence.
Protein level
binding fitness from KRAS-PIK3CGRDB bindingPCA of KRAS block1
The energetic and allosteric landscape for KRAS inhibition
Gene
KRAS
Score
-0.128 · p.Gly12Cys (assay numbering)
No calibration data
This score is at approximately the 82th percentile of all variants tested in this assay.
This score has not been calibrated against ACMG evidence — shown for context only, not as clinical evidence.
Protein level
binding fitness from KRAS-RALGDSRDB bindingPCA of KRAS block1
The energetic and allosteric landscape for KRAS inhibition
Gene
KRAS
Score
-0.0581 · p.Gly12Cys (assay numbering)
No calibration data
This score is at approximately the 88th percentile of all variants tested in this assay.
This score has not been calibrated against ACMG evidence — shown for context only, not as clinical evidence.
Protein level
binding fitness from KRAS-RAF1RDB bindingPCA of KRAS block1
The energetic and allosteric landscape for KRAS inhibition
Gene
KRAS
Score
-0.0926 · p.Gly12Cys (assay numbering)
No calibration data
This score is at approximately the 87th percentile of all variants tested in this assay.
This score has not been calibrated against ACMG evidence — shown for context only, not as clinical evidence.
Protein level
binding free energy changes of KRAS-SOS1 interaction
The energetic and allosteric landscape for KRAS inhibition
Gene
KRAS
Score
-0.166 · p.Gly12Cys (assay numbering)
No calibration data
This score is at approximately the 17th percentile of all variants tested in this assay.
This score has not been calibrated against ACMG evidence — shown for context only, not as clinical evidence.
Protein level
binding free energy changes of KRAS-RAF1RBD interaction
The energetic and allosteric landscape for KRAS inhibition
Gene
KRAS
Score
0.0397 · p.Gly12Cys (assay numbering)
No calibration data
This score is at approximately the 50th percentile of all variants tested in this assay.
This score has not been calibrated against ACMG evidence — shown for context only, not as clinical evidence.
Protein level
folding free energy changes of KRAS in abundancePCA
The energetic and allosteric landscape for KRAS inhibition
Gene
KRAS
Score
0.336 · p.Gly12Cys (assay numbering)
No calibration data
This score is at approximately the 49th percentile of all variants tested in this assay.
This score has not been calibrated against ACMG evidence — shown for context only, not as clinical evidence.
Protein level
binding fitness from KRAS-DARPin K27 bindingPCA of KRAS block1
The energetic and allosteric landscape for KRAS inhibition
Gene
KRAS
Score
-0.461 · p.Gly12Cys (assay numbering)
No calibration data
This score is at approximately the 61th percentile of all variants tested in this assay.
This score has not been calibrated against ACMG evidence — shown for context only, not as clinical evidence.
Protein level
binding free energy changes of KRAS-RALGDSRBD interaction
The energetic and allosteric landscape for KRAS inhibition
Gene
KRAS
Score
-0.102 · p.Gly12Cys (assay numbering)
No calibration data
This score is at approximately the 25th percentile of all variants tested in this assay.
This score has not been calibrated against ACMG evidence — shown for context only, not as clinical evidence.
Protein level
binding fitness from KRAS-DARPin K55 bindingPCA of KRAS block1
The energetic and allosteric landscape for KRAS inhibition
Gene
KRAS
Score
-0.110 · p.Gly12Cys (assay numbering)
No calibration data
This score is at approximately the 87th percentile of all variants tested in this assay.
This score has not been calibrated against ACMG evidence — shown for context only, not as clinical evidence.
Protein level
binding free energy changes of KRAS-DARPin K27 interaction
The energetic and allosteric landscape for KRAS inhibition
Gene
KRAS
Score
0.491 · p.Gly12Cys (assay numbering)
No calibration data
This score is at approximately the 84th percentile of all variants tested in this assay.
This score has not been calibrated against ACMG evidence — shown for context only, not as clinical evidence.
Protein level
binding free energy changes of KRAS-PIK3CGRBD interaction
The energetic and allosteric landscape for KRAS inhibition
Gene
KRAS
Score
0.0131 · p.Gly12Cys (assay numbering)
No calibration data
This score is at approximately the 40th percentile of all variants tested in this assay.
This score has not been calibrated against ACMG evidence — shown for context only, not as clinical evidence.
Protein level
binding free energy changes of KRAS-DARPin K55 interaction
The energetic and allosteric landscape for KRAS inhibition
Gene
KRAS
Score
0.0666 · p.Gly12Cys (assay numbering)
No calibration data
This score is at approximately the 55th percentile of all variants tested in this assay.
This score has not been calibrated against ACMG evidence — shown for context only, not as clinical evidence.

Publications

Other links and lift over

For research and educational, non-commercial use only. Not for clinical or diagnostic use. GeneBe does not provide medical advice. Data use for AI modeling is prohibited: if used, the cost is $0.001 per byte of downloaded uncompressed data.