FusionNeoAntigen Logo

Home

Download

Statistics

Examples

Help

Contact

Terms of Use

Center for Computational Systems Medicine
leaf

Fusion Gene and Fusion Protein Summary

leaf

Fusion Amino Acid Sequences (multiple BPs and multiple gene isoforms)

leaf

Fusion Protein Breakpoint Sequences - (for the Screening of the FusionNeoAntigens)

leaf

Potential FusionNeoAntigens in HLA I - (netMHCpan v4.1 + deepHLApan v1.1)

leaf

Potential FusionNeoAntigens in HLA II - (netMHCIIpan v4.1)

leaf

Fusion Breakpoint 14 AA Peptide Structure - (RoseTTAFold)

leaf

Filtering FusionNeoAntigens Through Checking the Interaction with HLAs in 3D - (Glide)

leaf

Vaccine Design for the FusionNeoAntigens (RNA/protein sequences)

leaf

Potential target of CAR-T therapy development

leaf

Information on the samples that have these potential fusion neoantigens

leaf

Fusion Protein Targeting Drugs - (Manual Curation)

leaf

Fusion Protein Related diseases - (Manual Curation)

Fusion Protein:PDK2-CDK12

Fusion Gene and Fusion Protein Summary

check button Fusion gene summary
Fusion partner gene informationFusion gene name: PDK2-CDK12
FusionPDB ID: 63934
FusionGDB2.0 ID: 63934
HgeneTgene
Gene symbol

PDK2

CDK12

Gene ID

5164

51755

Gene namepyruvate dehydrogenase kinase 2cyclin dependent kinase 12
SynonymsPDHK2|PDKIICRK7|CRKR|CRKRS
Cytomap

17q21.33

17q12

Type of geneprotein-codingprotein-coding
Descriptionpyruvate dehydrogenase kinase, isozyme 2PDH kinase 2[Pyruvate dehydrogenase (acetyl-transferring)] kinase isozyme 2, mitochondrialpyruvate dehydrogenase kinase, isoenzyme 2pyruvate dehydrogenase, lipoamide, kinase isozyme 2, mitochondrialcyclin-dependent kinase 12CDC2-related protein kinase 7Cdc2-related kinase, arginine/serine-richcell division cycle 2-related protein kinase 7cell division protein kinase 12
Modification date2020032020200313
UniProtAcc.

Q9NYV4

Main function of 5'-partner protein: FUNCTION: Cyclin-dependent kinase that phosphorylates the C-terminal domain (CTD) of the large subunit of RNA polymerase II (POLR2A), thereby acting as a key regulator of transcription elongation. Regulates the expression of genes involved in DNA repair and is required for the maintenance of genomic stability. Preferentially phosphorylates 'Ser-5' in CTD repeats that are already phosphorylated at 'Ser-7', but can also phosphorylate 'Ser-2'. Required for RNA splicing, possibly by phosphorylating SRSF1/SF2. Involved in regulation of MAP kinase activity, possibly leading to affect the response to estrogen inhibitors. {ECO:0000269|PubMed:11683387, ECO:0000269|PubMed:19651820, ECO:0000269|PubMed:20952539, ECO:0000269|PubMed:22012619, ECO:0000269|PubMed:24662513}.
Ensembl transtripts involved in fusion geneENST idsENST00000007708, ENST00000503176, 
ENST00000511026, 
ENST00000559545, 
ENST00000430627, ENST00000447079, 
Fusion gene scores for assessment (based on all fusion genes of FusionGDB 2.0)* DoF score3 X 4 X 2=2436 X 30 X 14=15120
# samples 555
** MAII scorelog2(5/24*10)=1.05889368905357
effective Gene in Pan-Cancer Fusion Genes (eGinPCFGs).
DoF>8 and MAII>0
log2(55/15120*10)=-4.78088271069641
possibly effective Gene in Pan-Cancer Fusion Genes (peGinPCFGs).
DoF>8 and MAII<0
Fusion gene context

PubMed: PDK2 [Title/Abstract] AND CDK12 [Title/Abstract] AND fusion [Title/Abstract]

Fusion neoantigen context

PubMed: PDK2 [Title/Abstract] AND CDK12 [Title/Abstract] AND neoantigen [Title/Abstract]

Most frequent breakpoint (based on all fusion genes of FusionGDB 2.0)PDK2(48182806)-CDK12(37627132), # samples:3
Anticipated loss of major functional domain due to fusion event.PDK2-CDK12 seems lost the major protein functional domain in Hgene partner, which is a CGC by not retaining the major functional domain in the partially deleted in-frame ORF.
PDK2-CDK12 seems lost the major protein functional domain in Hgene partner, which is a CGC by not retaining the major functional domain in the partially deleted in-frame ORF.
PDK2-CDK12 seems lost the major protein functional domain in Hgene partner, which is a essential gene by not retaining the major functional domain in the partially deleted in-frame ORF.
PDK2-CDK12 seems lost the major protein functional domain in Hgene partner, which is a essential gene by not retaining the major functional domain in the partially deleted in-frame ORF.
* DoF score (Degree of Frequency) = # partners X # break points X # cancer types
** MAII score (Major Active Isofusion Index) = log2(# samples/DoF score*10)

check button Gene ontology of each fusion partner gene with evidence of Inferred from Direct Assay (IDA) from Entrez
PartnerGeneGO IDGO termPubMed ID
TgeneCDK12

GO:0046777

protein autophosphorylation

11683387



check button Four levels of functional features of fusion genes
Go to FGviewer search page for the most frequent breakpoint (https://ccsmweb.uth.edu/FGviewer/chr17:48182806/chr17:37627132)
- FGviewer provides the online visualization of the retention search of the protein functional features across DNA, RNA, protein, and pathological levels.
- How to search
1. Put your fusion gene symbol.
2. Press the tab key until there will be shown the breakpoint information filled.
4. Go down and press 'Search' tab twice.
4. Go down to have the hyperlink of the search result.
5. Click the hyperlink.
6. See the FGviewer result for your fusion gene.
FGviewer

check buttonRetention analysis results of each fusion partner protein across 39 protein features of UniProt such as six molecule processing features, 13 region features, four site features, six amino acid modification features, two natural variation features, five experimental info features, and 3 secondary structure features, are available here.

check buttonFusion gene breakpoints across PDK2 (5'-gene)
* Click on the image to open the UCSC genome browser with custom track showing this image in a new window.
all structure

check buttonFusion gene breakpoints across CDK12 (3'-gene)
* Click on the image to open the UCSC genome browser with custom track showing this image in a new window.
all structure


Top

Fusion Amino Acid Sequences


check buttonFusion information from ORFfinder translation from full-length transcript sequence from FusionPDB.
HenstTenstHgeneHchrHbpHstrandTgeneTchrTbpTstrandSeq length
(transcript)
BP loci
(transcript)
Predicted start
(transcript)
Predicted stop
(transcript)
Seq length
(amino acids)
ENST00000007708PDK2chr1748182806+ENST00000430627CDK12chr1737627132+509253129539301211
ENST00000007708PDK2chr1748182806+ENST00000447079CDK12chr1737627132+778853129539571220
ENST00000503176PDK2chr1748182806+ENST00000430627CDK12chr1737627132+50544932038921290
ENST00000503176PDK2chr1748182806+ENST00000447079CDK12chr1737627132+77504932039191299
ENST00000007708PDK2chr1748182806+ENST00000430627CDK12chr1737627131+509253129539301211
ENST00000007708PDK2chr1748182806+ENST00000447079CDK12chr1737627131+778853129539571220
ENST00000503176PDK2chr1748182806+ENST00000430627CDK12chr1737627131+50544932038921290
ENST00000503176PDK2chr1748182806+ENST00000447079CDK12chr1737627131+77504932039191299

check buttonDeepORF prediction of the coding potential based on the fusion transcript sequence of in-frame fusion genes. DeepORF is a coding potential classifier based on convolutional neural network by comparing the real Ribo-seq data. If the no-coding score < 0.5 and coding score > 0.5, then the in-frame fusion transcript is predicted as being likely translated.
HenstTenstHgeneHchrHbpHstrandTgeneTchrTbpTstrandNo-coding scoreCoding score
ENST00000007708ENST00000430627PDK2chr1748182806+CDK12chr1737627132+0.0011606730.9988393
ENST00000007708ENST00000447079PDK2chr1748182806+CDK12chr1737627132+0.0001882640.9998117
ENST00000503176ENST00000430627PDK2chr1748182806+CDK12chr1737627132+0.0009178410.9990822
ENST00000503176ENST00000447079PDK2chr1748182806+CDK12chr1737627132+0.0001359470.999864
ENST00000007708ENST00000430627PDK2chr1748182806+CDK12chr1737627131+0.0011606730.9988393
ENST00000007708ENST00000447079PDK2chr1748182806+CDK12chr1737627131+0.0001882640.9998117
ENST00000503176ENST00000430627PDK2chr1748182806+CDK12chr1737627131+0.0009178410.9990822
ENST00000503176ENST00000447079PDK2chr1748182806+CDK12chr1737627131+0.0001359470.999864

check button Predicted full-length fusion amino acid sequences. For individual full-length fusion transcript sequence from FusionPDB, we ran ORFfinder and chose the longest ORF among all the predicted ones.

Get the fusion protein sequences from here.

Fusion protein sequence information is available in the fasta format.
>FusionGDB ID_FusionGDB isoform ID_FGname_Hgene_Hchr_Hbp_Henst_Tgene_Tchr_Tbp_Tenst_length(fusion AA) seq_BP

Top

Fusion Protein Breakpoint Sequences for PDK2-CDK12

check button +/-13 AA sequence from the breakpoints of the fusion protein sequences.
HgeneHchrHbpTgeneTchrTbpLength(fusion protein)BP in fusion proteinPeptide
PDK2chr1748182806CDK12chr1737627131493156EFLDKDPEDHRTLSHRKSMKSRSRSP
PDK2chr1748182806CDK12chr173762713153177EFLDKDPEDHRTLSHRKSMKSRSRSP
PDK2chr1748182806CDK12chr1737627132493156EFLDKDPEDHRTLSHRKSMKSRSRSP
PDK2chr1748182806CDK12chr173762713253177EFLDKDPEDHRTLSHRKSMKSRSRSP

Top

Potential FusionNeoAntigen Information of PDK2-CDK12 in HLA I

check button Multiple sequence alignments of the potential FusionNeoAntigens per fusion breakpoints. If the MSA is empty, then it means that there were predicted fusion neoantigens in this fusion breakpoint, but those predicted fusion neoantigens were not across the breakpoint, which is not fusion-specific.
PDK2-CDK12_48182806_37627131.msa

check button Potential FusionNeoAntigen Information
* We used NetMHCpan v4.1 (%rank<0.5) and deepHLApan v1.1 (immunogenic score>0.5)
Fusion geneHchrHbpTgeneTchrTbpHLA IFusionNeoAntigen peptideBinding scoreImmunogenic scoreNeoantigen start (at BP 13)Neoantigen end (at BP 13)
PDK2-CDK12chr1748182806chr1737627131531HLA-B15:17RTLSHRKSM0.95150.89411019
PDK2-CDK12chr1748182806chr1737627131531HLA-B15:16RTLSHRKSM0.92660.76141019
PDK2-CDK12chr1748182806chr1737627131531HLA-A30:08RTLSHRKSM0.91360.72291019
PDK2-CDK12chr1748182806chr1737627131531HLA-A30:08RTLSHRKSMK0.99760.75861020
PDK2-CDK12chr1748182806chr1737627131531HLA-C15:04RTLSHRKSM0.98130.92491019
PDK2-CDK12chr1748182806chr1737627131531HLA-C15:06RTLSHRKSM0.91280.93241019
PDK2-CDK12chr1748182806chr1737627131531HLA-C03:08RTLSHRKSM0.78330.89611019
PDK2-CDK12chr1748182806chr1737627131531HLA-C06:03RTLSHRKSM0.28130.99381019
PDK2-CDK12chr1748182806chr1737627131531HLA-C12:04RTLSHRKSM0.2780.99321019
PDK2-CDK12chr1748182806chr1737627131531HLA-C12:12RTLSHRKSM0.27230.94781019
PDK2-CDK12chr1748182806chr1737627131531HLA-C15:09RTLSHRKSM0.98130.92491019
PDK2-CDK12chr1748182806chr1737627131531HLA-A32:01RTLSHRKSM0.95360.95431019
PDK2-CDK12chr1748182806chr1737627131531HLA-B58:06RTLSHRKSM0.94240.79661019
PDK2-CDK12chr1748182806chr1737627131531HLA-A30:01RTLSHRKSM0.92620.90971019
PDK2-CDK12chr1748182806chr1737627131531HLA-C15:05RTLSHRKSM0.90670.90121019
PDK2-CDK12chr1748182806chr1737627131531HLA-C15:02RTLSHRKSM0.87310.89671019
PDK2-CDK12chr1748182806chr1737627131531HLA-C16:01RTLSHRKSM0.69330.98061019
PDK2-CDK12chr1748182806chr1737627131531HLA-C16:04RTLSHRKSM0.68440.97471019
PDK2-CDK12chr1748182806chr1737627131531HLA-C16:02RTLSHRKSM0.66290.99241019
PDK2-CDK12chr1748182806chr1737627131531HLA-C03:05RTLSHRKSM0.6160.86591019
PDK2-CDK12chr1748182806chr1737627131531HLA-C03:17RTLSHRKSM0.55290.95781019
PDK2-CDK12chr1748182806chr1737627131531HLA-C12:03RTLSHRKSM0.230.97881019
PDK2-CDK12chr1748182806chr1737627131531HLA-A30:01RTLSHRKSMK0.99780.90161020

Top

Potential FusionNeoAntigen Information of PDK2-CDK12 in HLA II

check button Multiple sequence alignments of the potential FusionNeoAntigens per fusion breakpoints. If the MSA is empty, then it means that there were predicted fusion neoantigens in this fusion breakpoint, but those predicted fusion neoantigens were not across the breakpoint, which is not fusion-specific.

check button Potential FusionNeoAntigen Information
* We used NetMHCIIpan v4.1 (%rank<0.5).
Fusion geneHchrHbpTgeneTchrTbpHLA IIFusionNeoAntigen peptideNeoantigen start (at BP 13)Neoantigen end (at BP 13)

Top

Fusion breakpoint peptide structures of PDK2-CDK12

check button3D structures of the fusion breakpoint peptide of 14AA sequence that have potential fusion neoantigens
* The minimum length of the amino acid sequence in RoseTTAFold is 14AA. Here, we predicted the 14AA fusion protein breakpoint sequence not the fusion neoantigen peptide, which is shorter than 14 AA.
File nameBPseqHgeneTgeneHchrHbpTchrTbpAAlen
6569PEDHRTLSHRKSMKPDK2CDK12chr1748182806chr1737627131531

Top

Filtering FusionNeoAntigens Through Checking the Interaction with HLAs in 3D of PDK2-CDK12

check buttonVirtual screening between 25 HLAs (from PDB) and FusionNeoAntigens
* We used Glide to predict the interaction between HLAs and neoantigens.
HLA allelePDB IDFile nameBPseqDocking scoreGlide score
HLA-B14:023BVN6569PEDHRTLSHRKSMK-7.15543-7.26883
HLA-B14:023BVN6569PEDHRTLSHRKSMK-4.77435-5.80965
HLA-B52:013W396569PEDHRTLSHRKSMK-6.80875-6.92215
HLA-B52:013W396569PEDHRTLSHRKSMK-4.20386-5.23916
HLA-A11:014UQ26569PEDHRTLSHRKSMK-7.5194-8.5547
HLA-A11:014UQ26569PEDHRTLSHRKSMK-6.9601-7.0735
HLA-A24:025HGA6569PEDHRTLSHRKSMK-7.52403-7.63743
HLA-A24:025HGA6569PEDHRTLSHRKSMK-5.82433-6.85963
HLA-B27:056PYJ6569PEDHRTLSHRKSMK-3.28285-4.31815
HLA-B44:053DX86569PEDHRTLSHRKSMK-5.91172-6.94702
HLA-B44:053DX86569PEDHRTLSHRKSMK-4.24346-4.35686

Top

Vaccine Design for the FusionNeoAntigens of PDK2-CDK12

check button mRNA and peptide sequences of FusionNeoAntigens that have potential interaction with HLA-Is.
Fusion geneHchrHbpTchrTbpStart in +/-13AAEnd in +/-13AAFusionNeoAntigen peptide sequenceFusionNeoAntigen RNA sequence
PDK2-CDK12chr1748182806chr17376271311019RTLSHRKSMCCTGAGCCATAGGAAATCCATGAAGTC
PDK2-CDK12chr1748182806chr17376271311020RTLSHRKSMKCCTGAGCCATAGGAAATCCATGAAGTCCAG

check button mRNA and peptide sequences of FusionNeoAntigens that have potential interaction with HLA-IIs.
Fusion geneHchrHbpTchrTbpStart in +/-13AAEnd in +/-13AAFusionNeoAntigen peptideFusionNEoAntigen RNA sequence

Top

Information of the samples that have these potential fusion neoantigens of PDK2-CDK12

check button These samples were reported as having these fusion breakpoints. For individual breakpoints, we checked the open reading frames considering multiple gene isoforms and chose the in-frame fusion genes only. Then, we made fusion protein sequences and predicted the fusion neoantigens. These fusion-positive samples may have these potential fusion neoantigens.
Cancer typeFusion geneHchrHbpHenstTchrTbpTenstSample
BRCAPDK2-CDK12chr1748182806ENST00000007708chr1737627131ENST00000430627TCGA-AR-A250

Top

Potential target of CAR-T therapy development for PDK2-CDK12

check button Predicted 3D structure. We used RoseTTAFold.

check buttonRetention analysis result of each fusion partner protein across 39 protein features of UniProt such as six molecule processing features, 13 region features, four site features, six amino acid modification features, two natural variation features, five experimental info features, and 3 secondary structure features. Here, to provide the retention of the transmembrane domain, we only show the protein feature retention information of those transmembrane features


* Minus value of BPloci means that the break point is located before the CDS.
- In-frame and retained 'Transmembrane'.
PartnerGeneHbpTbpENSTStrandBPexonTotalExonProtein feature loci*BPlociTotalLenProtein featureProtein feature note

check button Subcellular localization prediction of the transmembrane domain retained fusion proteins
* We used DeepLoc 1.0. The order of the X-axis of the barplot is as follows: Entry_ID, Localization, Type, Nucleus, Cytoplasm, Extracellular, Mitochondrion, Cell_membrane, Endoplasmic_reticulum, Plastid, Golgi.apparatus, Lysosome.Vacuole, Peroxisome. Y-axis is the output score of DeepLoc. Clicking the image will open a new tab with a large image.
HgeneHchrHbpHenstTgeneTchrTbpTenstDeepLoc result

Top

Related Drugs to PDK2-CDK12

check button Drugs used for this fusion-positive patient.
(Manual curation of PubMed, 04-30-2022 + MyCancerGenome)
HgeneTgeneDrugSourcePMID

Top

Related Diseases to PDK2-CDK12

check button Diseases that have this fusion gene.
(Manual curation of PubMed, 04-30-2022 + MyCancerGenome)
HgeneTgeneDiseaseSourcePMID

check button Diseases associated with fusion partners.
(DisGeNet 4.0)
PartnerGeneDisease IDDisease name# pubmedsSource