SUPPLEMENTARY MATERIAL Impact of knowledge accumulation on pathway enrichment analysis Lina Wadi1, Mona Meyer1, Joel Weiser1, Lincoln Stein1,3, Jüri Reimand1,2,*
1 - Ontario Institute for Cancer Research 2 - Department of Medical Biophysics, University Of Toronto 3 - Department of Molecular Genetics, University Of Toronto * - Correspondence: [email protected] Supplementary information, Wadi et al.
Supplementary Figures:...... 4
Supplementary Figure 1: The majority of web-based pathway analysis tools are out of date...... 5
Supplementary Figure 2: Vocabulary of biological pathways and processes is growing rapidly...... 6
Supplementary Figure 3: Gene Ontology terms are increasingly specific...... 7
Supplementary Figure 4: GO terms are increasingly interconnected...... 8
Supplementary Figure 5: Human genes and pathways are increasingly annotated...... 9
Supplementary Figure 6: Genes and pathways of model organisms are increasingly annotated...... 10
Supplementary Figure 7: Dark matter of protein-coding genome is decreasing.. 11
Supplementary Figure 8: Annotation analysis is affected by mismatches in gene symbols...... 12
Supplementary Figure 9: Enrichment analysis of breast cancer essential genes using gene annotations from 2010 misses numerous pathways and processes...... 13
Supplementary Figure 10: Pathway analysis with smaller gene lists confirms the large fraction of pathways missed in analysis of breast cancer genes...... 14
Supplementary Figure 11: Analysis of GO and Reactome pathways with breast cancer data confirms loss of information when using outdated software...... 15
Supplementary Figure 12: Evolution of pathway information affects recently updated and out-of-date software tools...... 16
2 Supplementary information, Wadi et al.
Supplementary Figure 13: Some enriched pathways and processes are missed in relatively recent gene annotations...... 17
Supplementary Figure 14: Most pathway enrichments from outdated annotations are based on low-quality information...... 18
Supplementary Tables...... 19
Supplementary Table 1: Number of citations of pathway analysis tools in 2015...... 20
Supplementary Table 2: Pathways only found in 2016 GO/ Reactome compared to 2010 pathway data...... 22
Supplementary Table 3a: GO terms 2016 compared to 2010...... 25
Supplementary Table 3b: Reactome terms 2016 compared to 2010...... 34
3 Supplementary information, Wadi et al.
Supplementary Figures:
4 Supplementary information, Wadi et al.
DAVID GeneCodis FuncAssociate PANTHER GeneTrail g:Profiler
EasyGO GoMiner gsGator Babelomics goEast
GARNet FunSpec WebGestalt GREAT ToppGene
ConceptGene BayGO goSurfer
L2L GOToolBox
2006 2008 2010 2012 2014 2016
Supplementary Figure 1: The majority of web-based pathway analysis tools are out of date. Shown are 21 pathway analysis tools and their time of update. Tools in the blue box have the latest updates prior to 2010.
5 Supplementary information, Wadi et al.
GO Biological Processes GO Cellular Components GO Molecular Functions Reactome Pathway 15000 4000 1500 1500 Year 2009 10000 3000 2010 1000 1000 2011 2012 2000 2013 5000 2014 500 500 1000 2015 2016
0 0 0 0
Number of terms with gene annotations 09 10 11 12 13 14 15 16 09 10 11 12 13 14 15 16 09 10 11 12 13 14 15 16 09 10 11 12 13 14 15 16
Supplementary Figure 2: Vocabulary of biological pathways and processes is growing rapidly. Comparison of Gene Ontology (GO) terms and Reactome pathways shows that the number of terms annotated to human genes has doubled in the period 2009-2016.
6 Supplementary information, Wadi et al.
20 Year 2009 2016
15
10 % of terms
5
0 2 3 4 5 6 7 8 9 10111213141516 Mean length of path to root per GO term
Supplementary Figure 3: Gene Ontology terms are increasingly specific. Histogram shows mean number of paths in the Gene Ontology connecting a given term and the root term. Significant increase in the depth of the GO hierarchy between 2009 and 2016 (p<10-5, permutation test) indicates that the biological vocabulary is increasingly detailed and terms are becoming more specific.
7 Supplementary information, Wadi et al.
Year 2009 2016 0.15
0.10 Density
0.05
0.00 0 5 10 Number of paths to root per GO term [log2]
Supplementary Figure 4: GO terms are increasingly interconnected. GO terms are connected to root term via significantly more paths in 2016 than in 2009 (p<10-5, permutation test), showing that biological concepts are increasingly interconnected. Complexity and depth of GO directly impacts enrichment analysis as gene annotations are hierarchically propagated.
8 Supplementary information, Wadi et al.
GO Reactome
p < 1x10-5 10 2022 2364
5 49 150 p < 1x10-5 Log2 median pathway size 0
10.0 p < 1x10-5 p = 0.00202
7.5 22 58
5.0
4 5 2.5 Log2 number of pathways 0.0 2009 2016 2009 2016 Year
Supplementary Figure 5: Human genes and pathways are increasingly annotated. Violin plots show median human pathway size (top plots) and the number of pathways per gene (bottom plots) in 2009 and 2016. P-values were generated using permutation tests with 100,000 permutations. Genes without annotations are excluded. GO biological processes (left) and Reactome (right) pathways are shown separately. The median number of each plot is shown in bold.
9 Supplementary information, Wadi et al.
Old New
Arabidopsis Fly Human Mouse Yeast -5 -5 -5 -5 -5 10.0 p < 1x10 p < 1x10 p < 1x10 p < 1x10 p < 1x10
7.5
5.0
2.5 Log2 number of pathways
0.0
Arabidopsis Fly Human Mouse Yeast 15
10
5 Log2 median pathway size p < 1x10-5 p < 1x10-5 p < 1x10-5 p < 1x10-5 p = 0.0048 0 2009 2016 2011 2016 2009 2016 2009 2016 2011 2016 Year
Supplementary Figure 6: Genes and pathways of model organisms are increasingly annotated. Violin plots show median pathway size (top plots) and the number of pathways per gene (bottom plots) in 2009 and 2016 for human and several model organisms. P-values were generated using permutation tests with 100,000 permutations. Genes without annotations are excluded.
10 Supplementary information, Wadi et al.
12
8
4
% of dark matter genes in CCDS 0 09 10 11 12 13 14 15 16
Year
Supplementary Figure 7: Dark matter of protein-coding genome is decreasing. The fraction of human genes with no annotations (i.e. dark matter) has increased consistently over time. Dark matter genes include human protein-coding genes from the most recent CCDS release that have no Reactome or GO annotations, or only have root-level GO annotations.
11 Supplementary information, Wadi et al.
12.5
10.0
7.5
5.0
2.5
0.0 2009 2011 2012 2013 2014 Gene symbols not found in 2015 CCDS (%) Year
Supplementary Figure 8: Annotation analysis is affected by mismatches in gene symbols. Use of out-dated annotations with current human gene lists will cause increasingly common mismatches of gene symbols as standard nomenclature is updated. Symbols of protein-coding genes from the latest (2015) release are compared to earlier releases of CCDS.
12 Supplementary information, Wadi et al.
cytokine pre-mrna excision nucleobase-containing dna-templated mediated ubiquitin-dependent fibroblast maintenance hiv transcription, tc-nermulti-organism heat polymerase cyclic canonical growth promoter proteins telomere macromolecule transcription expression stability transduction chromosome death amino decay g1/s aromatic formation membrane processing nucleusproteasome fgfr3degradation amide gene mrna nuclear-transcribed intronless organic mhc conjugation fgfr4 localization heterocycle targeting mitoticacid receptor integrity host dna cajal maturation transport subunit innate repair kinase virus activity necrosis telomerase metabolic
substance initiation shc1 ribosomal preinitiation cell p53 ribonucleoproteinexport 3'-end viral signaling modification ras antigen replication hormone stress immune nuclear rrna transition downstream metabolism peptide responserna nf-kb wnt assembly phosphorylation nitrogensignal translation complex stimulus exogenous arrest
cycle frs- nucleotide-excision termination phase damage insulin mapk activation transcript disassembly catabolic intracellular fgfr2 signalling mediator elongation checkpoint organization morphology cotranslational regulation biosynthetic synthesis amine biogenesis organelle ubiquitination translesion hedgehog (nmd) fgfr1single-organism organonitrogen erbb2 proteolysis egfr physiology independent activated nucleotideresponse, cascade
Supplementary Figure 9: Enrichment analysis of breast cancer essential genes using gene annotations from 2010 misses numerous pathways and processes. Word cloud shows most frequent keywords that are missed when analyzing essential breast cancer genes for enriched pathways and processes with out-dated functional annotations from 2010, compared to 2016 annotations. Text size corresponds to frequency of keywords from analysis of top-500 genes from shRNA screens of 77 cell lines.
13 Supplementary information, Wadi et al.
GO & Reactome: 2010 vs 2016 800 Type Both New
ms Old r
e 600 t
400
200 Number of significant
0 Cell line
Supplementary Figure 10: Pathway analysis with smaller gene lists confirms the large fraction of pathways missed in analysis of breast cancer genes. Pathway analysis with top-100 essential genes from 77 breast cancer cell lines ~three-fold enrichment of pathways and processes when data are analyzed with annotations from 2016 compared to 2010. GO biological processes and Reactome pathways are aggregated.
14 Supplementary information, Wadi et al.
A. GO: 2010 vs 2016 Type Both 600 New ms
r Old e t
400
200 Number of significant 0 ly2 vsat kpl1 mx1 l100 bt20 t47d jimt1 mcf7 e cal51 b skbr7 skbr5 skbr3 zr75b bt483 bt474 bt549 zr751 hcc70 hcc38 efm19 hdqp1 au565 sw527 cal148 cal851 cal120 h sum52 cama1 sum44 hs578t ocubm mb157 macls2 hcc202 mcf12a mcf10a du4475 X184b5 X184a1 sum225 sum229 sum149 sum102 sum185 sum159 sum190 mfm223 hcc1500 hcc1428 hcc1143 hcc1419 hcc1937 hcc1954 uacc893 hcc1806 hcc2185 hcc1187 hcc1599 hcc3153 hcc2218 hcc1569 hcc1008 hcc2688 hcc1395 efm192a sum1315 X600mpe mdamb361 mdamb453 mdamb468 mdamb436 mdamb231 mdamb330 mdamb415 mdamb157 mdamb134vi
B. Reactome: 2010 vs 2016
400 ms r e t 300
200
100 Number of significant 0 ly2 vsat kpl1 l100 mx1 bt20 t47d jimt1 mcf7 e cal51 b skbr3 skbr5 skbr7 zr751 zr75b bt549 bt483 bt474 hcc38 hcc70 au565 efm19 hdqp1 sw527 cal120 cal148 cal851 h sum44 ocubm hs578t sum52 cama1 mb157 macls2 hcc202 mcf10a mcf12a du4475 X184a1 X184b5 sum185 sum102 sum190 sum225 sum159 sum229 sum149 mfm223 hcc2185 hcc2688 hcc1569 hcc2218 hcc1008 hcc1599 hcc1395 hcc3153 hcc1187 hcc1428 hcc1143 hcc1500 hcc1954 hcc1419 uacc893 hcc1937 hcc1806 efm192a sum1315 X600mpe mdamb415 mdamb157 mdamb231 mdamb453 mdamb468 mdamb361 mdamb436 mdamb330 mdamb134vi Cell line
Supplementary Figure 11: Analysis of GO and Reactome pathways with breast cancer data confirms loss of information when using outdated software. Bar plots show the number of 2010-only, 2016-only and commonly detected enriched pathways for GO (top) and Reactome (bottom).
15 Supplementary information, Wadi et al.
Reactome terms 100
75
50
Proportion of terms (%) 25
0 2009 2010 2011 2012 2013 2014 2015 Year
Supplementary Figure 12: Evolution of pathway information affects recently updated and out-of-date software tools. Analysis of glioblastoma genes shows fraction of commonly detected (yellow), 2016-only (purple) and outdated-only (dark blue) Reactome pathways detected as significantly enriched. Significantly mutated glioblastoma genes were analyzed for pathway enrichments using annotations from 2009-2015 and compared to pathways detected using 2016 annotations.
16 Supplementary information, Wadi et al.
75
Category 50 Terms not present 0.05 < FDR < 0.1 0.1 < FDR < 0.2 FDR > 0.2
Missing terms in 2016 25
0
GO REACTOME
Supplementary Figure 13: Some enriched pathways and processes are missed in relatively recent gene annotations. In the 2015 analysis, 89/743 (12%) of GO terms and 29/116 (25%) Reactome pathways are missed compared to 2016.
17 Supplementary information, Wadi et al.
Development Apoptosis Heart Embryo Protein localization Tube
Behaviour Eye Sex Neuron CNS Axon Cell migration
Signalling Notch PI3K Steroid Insulin response FGFR VEGFR
MAPK RAS Glucose import/ transport NGF EGFR Histones & ERBB Immune signalling chromatin EGFRvIII Circardian PDGFR Wound clock healing
Response to stimulus Cellular Protein component modification organization Cell-cell signalling Cell cycle/ mitosis Kinase signalling
DNA Proliferation/ damage Adhesion stem cells response
Transcription Metabolism
GO 2016 Reactome 2016 GO 2010 Reactome 2010
Supplementary Figure 14: Most pathway enrichments from outdated annotations are based on low-quality information. Pathway enrichment analysis with only high-quality annotations misses 96.5% of results from 2016 analysis when 2010 annotations are used. The enrichment map confirms that the increase in enriched pathways is also observed when IEA annotations (Inferred from Electronic Annotations) are excluded from GO.
18 Supplementary information, Wadi et al.
Supplementary Tables
19 Supplementary information, Wadi et al.
Supplementary Table 1: Number of citations of pathway analysis tools in 2015. The table shows 21 different pathway tools, the number of citations per publication, the total number of citations of the tool, and the percentage of all citations in 2015.
Tool Citation count PMID Total citations in Total citations in 2015 per 2015 % publication
Babelomics 2 16845052 29 0.82 5 15980512 17 20478823 3 18515841 2 25897133 BayGO 0 16504085 0 0 ConceptGen 6 20007254 6 0.17
DAVID 1417 19131956 2517 71.42 261 12734009 614 19033363 28 14519205 72 17576678 57 17784955 16 17980028 21 19728287 23 22543366 8 18841237 EasyGO 3 17645808 3 0.09 FuncAssociate 13 19717575 21 0.60 8 14668247 FunSpec 16 12431279 16 0.45 g:Profiler1 26 17478515 74 2.10
gsGator 0 24423189 0 0.00
48 21646343
20 Supplementary information, Wadi et al.
Tool Citation count PMID Total citations in Total citations in 2015 per 2015 % publication
GARNet 0 21342555 0 0 GeneCodis 31 17204154 104 2.95 28 19465387 45 22573175
GeneTrail 12 17526521 14 0.40 2 21592396 GoEast 32 18487275 40 1.14 8 19615110
GoMiner 26 12702209 33 0.94 7 15998470 GoSurfer 1 15702958 1 0.03 GOToolBox 3 15575967 3 0.09
GREAT 142 20436461 145 4.11 3 23814184 L2L 0 16168088 0 0 PANTHER 16 19597783 242 6.87 8 17130144 132 23193289 86 23868073 ToppGene 72 19465376 72 2.04
Webgestalt 89 15980575 204 5.79 113 23703215 2 24233776
TOTAL 3524 3524
21 Supplementary information, Wadi et al.
Supplementary Table 2: Pathways only found in 2016 GO/ Reactome compared to 2010 pathway data. The table shows nine major themes/ pathways out of 28 that only come up when the data are analyzed with updated pathway tools. Similarly, new sub-pathways (spread across different major themes) occur with pathway tools updated in 2016. Pink: pathways only seen in 2016; yellow: pathways shared by 2010 and 2016.
Major themes/ pathways Sub-pathways
Catabolism Protein catabolism, RNA catabolism GABAergic synaptic Synaptic plasticity, neurotransmitter transport transmission Cognition/ learning/ behaviour Visual behaviour, associative learning, memory, cognition Glucose import/ transport Carbohydrate homeostasis, response to glucose stimulus Circadian clock Regulation of circardian rhythm, BMAL1:CLOCK:NPAS2 activates circardian gene expression Wound healing Coagulation, platelet activation, homeostasis Immune signalling Fc receptor signalling pathway, TCR signalling, cytokine signalling Homeostasis Chemical homeostasis, tissue homeostasis Endocytosis Vesicle-mediated transport, receptor-mediated endocytosis Adhesion Cell-matrix adhesion, cell-substrate adhesion Response to stimulus Response to cAMP/ purine-containing compound Response to EGF Response to UV Response to stress Response to metal ion/ inorganic substances Response to TGFβ Cellular component RNP complex biogenesis/ assembly organization Negative regulation of organelle / cellular component organization
22 Supplementary information, Wadi et al.
Major themes/ pathways Sub-pathways Histones & chromatin H3-K9 methylation Histone acetylation/ peptidyl-lysine acetylation Chromatin (dis)assembly DNA alkylation/ (de)methylation Histone H3 acetylation Histone deacetylation Peptidyl-lysine acetylation Nucleosome organization Histone H3-K9 acetylation Chromatin modifying enzymes DNA replication/ CC checkpoint Regulation of DNA replication Cell cycle/ mitosis mitotic CC phase transition Chromosome segregation/ meta-anaphase transition Sister chromatid cohesion Sister chromatid segregation Apoptosis/ Neuron death Neuron death regulation Chromosome breakage/ programmed DNA elimination Fibroblast apoptotic process Neuron apoptotic process Cell death signalling via NRAGE/ NRIF/ NADE Mesenchymal apoptotic process p75 NTR receptor-mediated signalling Cell type specific apoptotic process Protein import/ localization Protein localization to membrane ECM organization Potassium ion transmembrane transport
23 Supplementary information, Wadi et al.
Major themes/ pathways Sub-pathways protein import into nucleus Metabolism Regulation of ROS Regulation of TNF Signalling EGFRvIII FGFR Notch VEGF ERK1 and ERK2 cascade ERBB TGFβ Development Tube Neuro/ axono/ gliogenesis Cartilage Head/ body/ face Trachea Eye Sertoli cell Liver Pattern specification
24 Supplementary information, Wadi et al.
Supplementary Table 3a: GO terms 2016 compared to 2010. The table shows the top most significant GO terms in 2016 ranked by FDR adjusted p-values compared to FDR in 2010.
FDR of GO.ID in GO.ID Description FDR Common genes 2010 AKAP9,ANK3,ARFGEF2,ARHG AP35,ARHGEF6,ARID1A,ATR X,BPTF,BRAF,BRCA1,CASP1,C HD8,CLOCK,CNOT1,CUL1,DIS 3,EGFR,EZH2,FN1,HSP90AB1 posi ve regula on ,KALRN,KDM6A,KDR,KRAS,L GO: of metabolic 2.69E-14 RP6,MAP3K4,MAP4K3,MEN 0.005040439 0009893 process 1,MET,NCOR1,NEDD4L,NF1, NFATC4,NR2F2,PAX5,PIK3CA, PIK3CB,PIK3R1,PTEN,PTPN11 ,RB1,SF3B1,SIN3A,SOS1,SOX 9,SPTAN1,STAG2,TGFBR2,TP5 3,TRIO,WT1,KMT2A AKAP9,ANK3,ARID1A,ATRX,B PTF,BRAF,BRCA1,CASP1,CHD 8,CLOCK,CNOT1,CUL1,EGFR, EZH2,FN1,HSP90AB1,KDM6A posi ve regula on GO: ,KDR,KRAS,LRP6,MAP3K4,M of macromolecule 2.61E-13 AP4K3,MEN1,MET,NCOR1,N 0.024811057 0010604 metabolic process EDD4L,NF1,NFATC4,NR2F2,P AX5,PIK3CA,PIK3CB,PIK3R1,P TEN,PTPN11,RB1,SF3B1,SIN3 A,SOS1,SOX9,SPTAN1,STAG2, TGFBR2,TP53,WT1,KMT2A ADAM10,AKAP9,ANK3,ARHG EF6,ATRX,BPTF,BRAF,BRCA1, CAD,CASP1,CUL1,EGFR,EZH2 ,FN1,HDAC9,HSP90AB1,KALR N,KDR,KRAS,LRP6,MAX,MEN GO: cellular response to 5.20E-13 1,MET,NCOR1,NEDD4L,NF1, ND 0070887 chemical s mulus NFATC4,NR2F2,NUP107,PIK3 CA,PIK3CB,PIK3R1,PRPF8,PT EN,PTPN11,RB1,SIN3A,SOS1, SOX9,SPTAN1,TGFBR2,TP53,T RIO,WT1,KMT2A
25 Supplementary information, Wadi et al.
FDR of GO.ID in GO.ID Description FDR Common genes 2010 ADAM10,AKAP9,ANK3,ARHG AP35,ARID1A,ATRX,BPTF,BR AF,BRCA1,CAD,CARM1,CHD8 ,CLOCK,CNOT1,CSDE1,CUL1, EGFR,EZH2,FAT1,FN1,HDAC9 anatomical ,HSP90AB1,IDH1,KALRN,KD GO: M6A,KDR,KRAS,LRP6,MAP3K structure 2.30E-12 0.002983586 0048856 4,MAX,MEN1,MET,NCOR1,N development EDD4L,NF1,NFATC4,NR2F2,P AX5,PBRM1,PCDH18,PIK3CA ,PIK3CB,PIK3R1,PTEN,PTPN1 1,RB1,SF3B1,SIN3A,SOS1,SO X9,SPTAN1,TGFBR2,TJP1,TP5 3,TRIO,WT1,RPSA,KMT2A ACAD8,AKAP9,ANK3,ARFGEF 2,ARHGAP35,ARHGEF6,ARID 1A,ARID2,ATRX,BPTF,BRAF,B RCA1,CARM1,CASP1,CHD8,C LOCK,CLTC,CNOT1,CSDE1,CU L1,DIS3,EGFR,EZH2,FN1,HDA C9,HSP90AB1,IDH1,KALRN,K GO: regula on of 2.70E-12 DM5C,KDM6A,KDR,KRAS,LR 0.000413789 0019222 metabolic process P6,MAP3K4,MAP4K3,MAX, MEN1,MET,NCOR1,NEDD4L, NF1,NFATC4,NR2F2,NUP107, PAX5,PBRM1,PIK3CA,PIK3CB ,PIK3R1,PTEN,PTPN11,RB1,S F3B1,SIN3A,SOS1,SOX9,SPTA N1,STAG2,TGFBR2,TP53,TRIO ,WT1,ZNF814,KMT2A ADAM10,AKAP9,ANK3,ARFG EF2,ARHGAP35,ARHGEF6,AR ID1A,ARID2,ATRX,BAP1,BPTF ,BRAF,BRCA1,CARM1,CHD8, CLOCK,CLTC,CNOT1,DIS3,EGF R,EZH2,FAT1,FN1,HDAC9,HS cellular component GO: P90AB1,KALRN,KDM5C,KDM organiza on or 2.75E-12 6A,KDR,KRAS,LRP6,MAP3K4, ND 0071840 biogenesis MAX,MEN1,MET,NCOR1,NE DD4L,NF1,NFATC4,NUP107,P AX5,PBRM1,PIK3CA,PIK3CB, PIK3R1,PRPF8,PTEN,PTPN11, RB1,RPL5,SF3B1,SIN3A,SOS1 ,SOX9,SPTAN1,STAG2,TJP1,TP 53,TRIO,WT1,RPSA,KMT2A
26 Supplementary information, Wadi et al.
FDR of GO.ID in GO.ID Description FDR Common genes 2010 ACAD8,ADAM10,AKAP9,ANK 3,AQR,ARFGEF2,ARHGAP35, ARHGEF6,ARID1A,ARID2,ATR X,BAP1,BPTF,BRAF,BRCA1,CA D,CARM1,CASP1,CHD8,CLOC K,CLTC,CNOT1,CSDE1,CUL1, DIS3,EGFR,EZH2,FN1,HDAC9, cellular HSP90AB1,KALRN,KDM5C,K GO: DM6A,KDR,KRAS,LRP6,MAP macromolecule 2.75E-12 3.94E-06 0044260 3K4,MAP4K3,MAX,MEN1,M metabolic process ET,NCOR1,NEDD4L,NF1,NFAT C4,NR2F2,NUP107,PAX5,PBR M1,PIK3CA,PIK3CB,PIK3R1,P RPF8,PTEN,PTPN11,RB1,RPL 5,SF3B1,SIN3A,SOS1,SOX9,S PTAN1,STAG2,TGFBR2,TP53,T RIO,WT1,ZNF814,RPSA,KMT 2A ADAM10,AKAP9,ANK3,ARHG AP35,ARID1A,ATRX,BPTF,BR AF,BRCA1,CAD,CARM1,CHD8 ,CLOCK,CNOT1,CSDE1,CUL1, EGFR,EZH2,FN1,HDAC9,HSP9 mul cellular 0AB1,IDH1,KALRN,KDM6A,K GO: organismal 5.30E-12 DR,KRAS,LRP6,MAP3K4,MAX 0.000751915 0007275 ,MEN1,MET,NCOR1,NEDD4L, development NF1,NFATC4,NR2F2,PAX5,PB RM1,PCDH18,PIK3CA,PIK3CB ,PIK3R1,PTEN,PTPN11,RB1,S F3B1,SIN3A,SOS1,SOX9,SPTA N1,TGFBR2,TJP1,TP53,TRIO, WT1,KMT2A ADAM10,AKAP9,ANK3,ARFG EF2,ARHGAP35,ARHGEF6,AR ID1A,ARID2,ATRX,BAP1,BPTF ,BRAF,BRCA1,CARM1,CHD8, CLOCK,CLTC,CNOT1,EGFR,EZ H2,FAT1,FN1,HDAC9,HSP90A GO: cellular component B1,KALRN,KDM5C,KDM6A,K 5.30E-12 DR,KRAS,LRP6,MAP3K4,MAX 1.60E+04 0016043 organiza on ,MEN1,MET,NCOR1,NEDD4L, NF1,NFATC4,NUP107,PAX5,P BRM1,PIK3CA,PIK3CB,PIK3R 1,PRPF8,PTEN,PTPN11,RB1,R PL5,SF3B1,SIN3A,SOS1,SOX9 ,SPTAN1,STAG2,TJP1,TP53,TR IO,WT1,RPSA,KMT2A
27 Supplementary information, Wadi et al.
FDR of GO.ID in GO.ID Description FDR Common genes 2010 ADAM10,AKAP9,ANK3,ARHG AP35,ARID1A,ATRX,BPTF,BR AF,BRCA1,CAD,CARM1,CHD8 ,CLOCK,CSDE1,CUL1,EGFR,EZ H2,FN1,HDAC9,HSP90AB1,ID H1,KALRN,KDM6A,KDR,KRAS GO: system 6.03E-12 ,LRP6,MAP3K4,MAX,MEN1, 0.005167947 0048731 development MET,NCOR1,NEDD4L,NF1,NF ATC4,NR2F2,PAX5,PBRM1,PC DH18,PIK3CA,PIK3CB,PIK3R1 ,PTEN,PTPN11,RB1,SIN3A,SO S1,SOX9,SPTAN1,TGFBR2,TP5 3,TRIO,WT1,KMT2A ADAM10,AKAP9,ANK3,ARHG AP35,ARID1A,ATRX,BPTF,BR AF,BRCA1,CAD,CARM1,CHD8 ,CLOCK,CLTC,CNOT1,CSDE1,C UL1,EGFR,EZH2,FAT1,FN1,HD AC9,HSP90AB1,IDH1,KALRN, GO: developmental KDM6A,KDR,KRAS,LRP6,MA 6.07E-12 P3K4,MAX,MEN1,MET,NCOR 0.000751915 0032502 process 1,NEDD4L,NF1,NFATC4,NR2F 2,PAX5,PBRM1,PCDH18,PIK3 CA,PIK3CB,PIK3R1,PTEN,PTP N11,RB1,SF3B1,SIN3A,SOS1, SOX9,SPTAN1,STAG2,TGFBR2 ,TJP1,TP53,TRIO,WT1,RPSA,K MT2A ARID1A,ARID2,ATRX,BAP1,B PTF,BRCA1,CARM1,CHD8,CL GO: chroma n 8.74E-12 OCK,EZH2,HDAC9,KDM5C,K 1.06E-07 0016568 modifica on DM6A,MEN1,NCOR1,PAX5,P BRM1,RB1,SIN3A,SOX9,TP53 ,KMT2A AKAP9,ARHGAP35,ARID1A,A TRX,BPTF,BRAF,BRCA1,CAD,C ARM1,CHD8,CLOCK,CSDE1,C UL1,EGFR,EZH2,HDAC9,HSP9 0AB1,IDH1,KDM6A,KDR,KRA GO: organ development 1.03E-11 S,LRP6,MAP3K4,MAX,MEN1, 0.000800217 0048513 MET,NCOR1,NF1,NFATC4,NR 2F2,PAX5,PBRM1,PCDH18,PI K3CA,PIK3R1,PTEN,PTPN11, RB1,SIN3A,SOS1,SOX9,TGFB R2,TP53,WT1,KMT2A
28 Supplementary information, Wadi et al.
FDR of GO.ID in GO.ID Description FDR Common genes 2010 ACAD8,AKAP9,ANK3,ARHGA P35,ARID1A,ARID2,ATRX,BPT F,BRAF,BRCA1,CARM1,CASP1 ,CHD8,CLOCK,CLTC,CNOT1,C SDE1,CUL1,DIS3,EGFR,EZH2, FN1,HDAC9,HSP90AB1,KDM regula on of 5C,KDM6A,KDR,KRAS,LRP6, GO: macromolecule 1.03E-11 MAP3K4,MAP4K3,MAX,MEN 0.002467951 0060255 metabolic process 1,MET,NCOR1,NEDD4L,NF1, NFATC4,NR2F2,NUP107,PAX 5,PBRM1,PIK3CA,PIK3CB,PIK 3R1,PTEN,PTPN11,RB1,SF3B 1,SIN3A,SOS1,SOX9,SPTAN1, STAG2,TGFBR2,TP53,WT1,ZN F814,KMT2A ADAM10,AKAP9,ANK3,ARHG AP35,ARID1A,ATRX,BPTF,BR AF,BRCA1,CAD,CARM1,CHD8 ,CLOCK,CLTC,CNOT1,CSDE1,C UL1,EGFR,EZH2,FN1,HDAC9, single-organism HSP90AB1,IDH1,KALRN,KDM GO: developmental 1.46E-11 6A,KDR,KRAS,LRP6,MAP3K4, ND 0044767 MAX,MEN1,MET,NCOR1,NE process DD4L,NF1,NFATC4,NR2F2,PA X5,PBRM1,PCDH18,PIK3CA,P IK3CB,PIK3R1,PTEN,PTPN11, RB1,SF3B1,SIN3A,SOS1,SOX9 ,SPTAN1,STAG2,TGFBR2,TJP1, TP53,TRIO,WT1,RPSA,KMT2A ADAM10,AKAP9,ARHGEF6,A TRX,BPTF,BRAF,BRCA1,CAD,C ARM1,CASP1,CUL1,EGFR,EZ H2,FN1,HDAC9,HSP90AB1,ID GO: response to organic H1,KALRN,KDR,KRAS,LRP6,M 3.16E-11 AP4K3,MAX,MEN1,NCOR1,N 0.032600063 0010033 substance EDD4L,NF1,NR2F2,NUP107,P IK3CA,PIK3CB,PIK3R1,PRPF8, PTEN,PTPN11,SIN3A,SOS1,S OX9,SPTAN1,TGFBR2,TP53,TR IO,WT1
29 Supplementary information, Wadi et al.
FDR of GO.ID in GO.ID Description FDR Common genes 2010 ADAM10,AKAP9,ANK3,ARHG AP35,ARID1A,ATRX,BRAF,BR CA1,CARM1,CUL1,EGFR,EZH anatomical 2,FAT1,FN1,HDAC9,HSP90AB GO: 1,KALRN,KDM6A,KDR,KRAS,L structure 8.42E-11 0.00126693 0009653 RP6,MET,NEDD4L,NF1,NFATC morphogenesis 4,NR2F2,PAX5,PIK3CA,PIK3C B,PTEN,PTPN11,RB1,SF3B1,S OS1,SOX9,SPTAN1,TGFBR2,TJ P1,TP53,TRIO,WT1 AKAP9,ARID1A,ATRX,BPTF,B RAF,BRCA1,CASP1,CHD8,CLO CK,CNOT1,CUL1,EGFR,EZH2, posi ve regula on FN1,HSP90AB1,KDR,KRAS,LR GO: of cellular metabolic 8.42E-11 P6,MAP3K4,MAP4K3,MEN1, 0.003630204 0031325 MET,NCOR1,NF1,NFATC4,NR process 2F2,PAX5,PIK3CA,PIK3CB,PIK 3R1,PTEN,PTPN11,RB1,SIN3 A,SOS1,SOX9,SPTAN1,STAG2, TGFBR2,TP53,WT1,KMT2A AKAP9,ARID1A,ATRX,BPTF,CS reproduc ve GO: DE1,EGFR,HSP90AB1,IDH1,K structure 8.42E-11 DR,LRP6,MAP3K4,MEN1,ME 0.015434574 0048608 development T,NR2F2,PBRM1,PTEN,PTPN 11,SOX9,WT1 AKAP9,ARHGEF6,BPTF,BRAF, cellular response to CAD,EGFR,FN1,KALRN,KDR,K GO: RAS,MEN1,NCOR1,NEDD4L, growth factor 8.42E-11 ND 0071363 NF1,PIK3CA,PIK3CB,PIK3R1,P s mulus TEN,PTPN11,SOS1,SOX9,SPT AN1,TGFBR2,TP53,TRIO AKAP9,ARID1A,ATRX,BPTF,CS GO: reproduc ve system DE1,EGFR,HSP90AB1,IDH1,K 8.63E-11 DR,LRP6,MAP3K4,MEN1,ME ND 0061458 development T,NR2F2,PBRM1,PTEN,PTPN 11,SOX9,WT1
30 Supplementary information, Wadi et al.
FDR of GO.ID in GO.ID Description FDR Common genes 2010 ACAD8,ADAM10,AKAP9,ANK 3,AQR,ARFGEF2,ARHGAP35, ARHGEF6,ARID1A,ARID2,ATR X,BAP1,BPTF,BRAF,BRCA1,CA D,CARM1,CASP1,CHD8,CLOC K,CLTC,CNOT1,CSDE1,CUL1, DIS3,EGFR,EZH2,FN1,HDAC9, HSP90AB1,KALRN,KDM5C,K GO: macromolecule DM6A,KDR,KRAS,LRP6,MAP 9.74E-11 2.36E-05 0043170 metabolic process 3K4,MAP4K3,MAX,MEN1,M ET,NCOR1,NEDD4L,NF1,NFAT C4,NR2F2,NUP107,PAX5,PBR M1,PIK3CA,PIK3CB,PIK3R1,P RPF8,PTEN,PTPN11,RB1,RPL 5,SF3B1,SIN3A,SOS1,SOX9,S PTAN1,STAG2,TGFBR2,TP53,T RIO,WT1,ZNF814,RPSA,KMT 2A ADAM10,AKAP9,ANK3,ARHG AP35,ARID1A,ATRX,BPTF,BR AF,BRCA1,CARM1,CHD8,EGF R,EZH2,FN1,HDAC9,HSP90A GO: nervous system 1.08E-10 B1,KALRN,KDM6A,KRAS,LRP ND 0007399 development 6,MEN1,MET,NCOR1,NEDD4 L,NF1,NFATC4,NR2F2,PAX5,P CDH18,PTEN,PTPN11,RB1,S OS1,SOX9,SPTAN1,TGFBR2,T P53,TRIO ACAD8,ADAM10,ANK3,AQR, ARHGAP35,ARID1A,ARID2,A TRX,BPTF,BRAF,BRCA1,CARM 1,CASP1,CHD8,CLOCK,CNOT 1,CSDE1,DIS3,EGFR,EZH2,FN 1,HDAC9,KDM5C,KDM6A,KR GO: gene expression 1.23E-10 AS,LRP6,MAP3K4,MAX,MEN 0.000703641 0010467 1,MET,NCOR1,NEDD4L,NF1, NFATC4,NR2F2,NUP107,PAX 5,PBRM1,PIK3CA,PIK3CB,PIK 3R1,PRPF8,PTEN,RB1,RPL5,S F3B1,SIN3A,SOX9,STAG2,TGF BR2,TP53,WT1,ZNF814,RPSA ,KMT2A
31 Supplementary information, Wadi et al.
FDR of GO.ID in GO.ID Description FDR Common genes 2010 AKAP9,ARHGEF6,BPTF,BRAF, CAD,EGFR,FN1,KALRN,KDR,K GO: response to growth RAS,MEN1,NCOR1,NEDD4L, 1.23E-10 ND 0070848 factor NF1,PIK3CA,PIK3CB,PIK3R1,P TEN,PTPN11,SOS1,SOX9,SPT AN1,TGFBR2,TP53,TRIO ADAM10,AKAP9,ANK3,ARFG EF2,ARHGAP35,ARHGEF6,AR ID1A,ATRX,BAP1,BPTF,BRAF, BRCA1,CARM1,CASP1,CHD8, CLOCK,CNOT1,CUL1,DIS3,EG FR,EZH2,FN1,HDAC9,HSP90A GO: posi ve regula on 1.23E-10 B1,KALRN,KDM6A,KDR,KRAS 3.11E-05 0048518 of biological process ,LRP6,MAP3K4,MAP4K3,ME N1,MET,NCOR1,NEDD4L,NF1 ,NFATC4,NR2F2,PAX5,PIK3CA ,PIK3CB,PIK3R1,PTEN,PTPN1 1,RB1,SF3B1,SIN3A,SOS1,SO X9,SPTAN1,STAG2,TGFBR2,TP 53,TRIO,WT1,KMT2A ADAM10,AKAP9,ANK3,ARHG AP35,ARHGEF6,ARID1A,ATR X,BAP1,BPTF,BRAF,BRCA1,CA RM1,CASP1,CHD8,CLOCK,CN OT1,CUL1,EGFR,EZH2,FN1,H GO: posi ve regula on DAC9,HSP90AB1,KALRN,KDR 1.61E-10 ,KRAS,LRP6,MAP3K4,MAP4K 5.84E-06 0048522 of cellular process 3,MEN1,MET,NCOR1,NEDD4 L,NF1,NFATC4,NR2F2,PAX5,P IK3CA,PIK3CB,PIK3R1,PTEN,P TPN11,RB1,SIN3A,SOS1,SOX 9,SPTAN1,STAG2,TGFBR2,TP5 3,TRIO,WT1,KMT2A ARID1A,ARID2,ATRX,BAP1,B PTF,BRCA1,CARM1,CHD8,CL OCK,EZH2,HDAC9,KDM5C,K GO: chromosome 1.76E-10 DM6A,MAP3K4,MEN1,NCOR 1.34E-06 0051276 organiza on 1,NUP107,PAX5,PBRM1,PTE N,RB1,SIN3A,SOX9,STAG2,TP 53,KMT2A ARID1A,ARID2,ATRX,BAP1,B PTF,BRCA1,CARM1,CHD8,CL GO: chroma n 1.88E-10 OCK,EZH2,HDAC9,KDM5C,K 1.95E-06 0006325 organiza on DM6A,MEN1,NCOR1,PAX5,P BRM1,RB1,SIN3A,SOX9,TP53 ,KMT2A
32 Supplementary information, Wadi et al.
FDR of GO.ID in GO.ID Description FDR Common genes 2010 ACAD8,ANK3,ARHGAP35,ARI D1A,ARID2,ATRX,BPTF,BRAF, BRCA1,CARM1,CHD8,CLOCK, CNOT1,CSDE1,DIS3,EGFR,EZ H2,FN1,HDAC9,KDM5C,KDM GO: regula on of gene 6A,KRAS,LRP6,MAP3K4,MAX 1.89E-10 0.004212933 0010468 expression ,MEN1,MET,NCOR1,NEDD4L, NF1,NFATC4,NR2F2,NUP107, PAX5,PBRM1,PIK3CA,PIK3CB ,PIK3R1,PTEN,RB1,SF3B1,SIN 3A,SOX9,STAG2,TGFBR2,TP5 3,WT1,ZNF814,KMT2A
33 Supplementary information, Wadi et al.
Supplementary Table 3b: Reactome terms 2016 compared to 2010. The table shows the top most significant Reactome terms in 2016 ranked by FDR adjusted p- values compared to FDR in 2010.
FDR of React.ID Description FDR Common genes Reactome.ID in 2010 ADAM10,AKAP9,ANK3,ARHGAP 35,BRAF,CARM1,CLTC,EGFR,EZ Developmental H2,FN1,HSP90AB1,KALRN,KDM R-HSA-1266738 3.44E-06 ND Biology 6A,KDR,KRAS,MET,NCOR1,NF1, NR2F2,PTPN11,SOS1,SPTAN1,T RIO R-HSA-210993 Tie2 Signaling 3.44E-06 KRAS,PIK3CA,PIK3CB,PIK3R1,PT 1.26E-05 PN11,SOS1 AKAP9,BRAF,CUL1,EGFR,FN1,K Signaling by RAS,NCOR1,NF1,PIK3CA,PIK3CB R-HSA-1236394 3.54E-06 ND ERBB4 ,PIK3R1,PTEN,PTPN11,SOS1,SP TAN1 ADAM10,AKAP9,ANK3,ARHGAP 35,BRAF,CLTC,EGFR,FN1,HSP90 R-HSA-422475 Axon guidance 9.12E-06 ND AB1,KALRN,KDR,KRAS,MET,NF1 ,PTPN11,SOS1,SPTAN1,TRIO Cons tu ve R-HSA-5637810 Signaling by 1.58E-05 EGFR,KRAS,PIK3CA,PIK3R1,SOS ND 1 EGFRvIII Signaling by EGFR,KRAS,PIK3CA,PIK3R1,SOS R-HSA-5637812 EGFRvIII in 1.58E-05 ND 1 Cancer Diseases of ADAM10,CUL1,EGFR,HDAC9,KR R-HSA-5663202 signal 1.67E-05 AS,LRP6,NCOR1,PIK3CA,PIK3CB ND transduc on ,PIK3R1,PTPN11,SOS1,TGFBR2 AKAP9,ARHGEF6,BRAF,EGFR,FN Signalling by R-HSA-166520 1.67E-05 1,KALRN,KRAS,NF1,PIK3CA,PIK 0.00410468 NGF 3CB,PIK3R1,PTEN,PTPN11,SOS1 ,SPTAN1,TRIO AKAP9,BRAF,CASP1,CUL1,EGFR, Signaling by R-HSA-449147 1.76E-05 FN1,KRAS,NF1,PIK3CA,PIK3CB, ND Interleukins PIK3R1,PTPN11,SOS1,SPTAN1 ADAM10,AKAP9,BRAF,EGFR,FN Signaling by R-HSA-177929 2.31E-05 1,KRAS,NF1,PIK3CA,PIK3CB,PIK 1.26E-05 EGFR 3R1,PTEN,PTPN11,SOS1,SPTAN 1
34 Supplementary information, Wadi et al.
FDR of React.ID Description FDR Common genes Reactome.ID in 2010 Cons tu ve Signaling by Ligand- EGFR,KRAS,PIK3CA,PIK3R1,SOS R-HSA-1236382 2.39E-05 ND Responsive 1 EGFR Cancer Variants ADAM10,CUL1,EGFR,HDAC9,HS P90AB1,IDH1,KRAS,LRP6,NCOR R-HSA-1643685 Disease 2.39E-05 1,NEDD4L,NUP107,PIK3CA,PIK3 ND CB,PIK3R1,PTPN11,RPL5,SOS1,T GFBR2,RPSA Fc epsilon AKAP9,BRAF,CUL1,EGFR,FN1,K R-HSA-2454202 receptor (FCERI) 2.39E-05 RAS,NF1,PIK3CA,PIK3CB,PIK3R1 ND signaling ,PTEN,PTPN11,SOS1,SPTAN1 Signaling by EGFR,KRAS,PIK3CA,PIK3R1,SOS R-HSA-1643713 2.39E-05 ND EGFR in Cancer 1 Signaling by Ligand- R-HSA-5637815 Responsive 2.39E-05 EGFR,KRAS,PIK3CA,PIK3R1,SOS ND 1 EGFR Variants in Cancer Downstream AKAP9,BRAF,EGFR,FN1,KRAS,N R-HSA-5654687 signaling of 2.97E-05 F1,PIK3CA,PIK3CB,PIK3R1,PTEN ND ac vated FGFR1 ,PTPN11,SOS1,SPTAN1 Downstream AKAP9,BRAF,EGFR,FN1,KRAS,N R-HSA-5654696 signaling of 2.97E-05 F1,PIK3CA,PIK3CB,PIK3R1,PTEN ND ac vated FGFR2 ,PTPN11,SOS1,SPTAN1 Downstream AKAP9,BRAF,EGFR,FN1,KRAS,N R-HSA-5654708 signaling of 2.97E-05 F1,PIK3CA,PIK3CB,PIK3R1,PTEN ND ac vated FGFR3 ,PTPN11,SOS1,SPTAN1 Downstream AKAP9,BRAF,EGFR,FN1,KRAS,N R-HSA-5654716 signaling of 2.97E-05 F1,PIK3CA,PIK3CB,PIK3R1,PTEN ND ac vated FGFR4 ,PTPN11,SOS1,SPTAN1 Interleukin-3, 5 AKAP9,BRAF,EGFR,FN1,KRAS,N R-HSA-512988 and GM-CSF 2.97E-05 F1,PIK3CA,PIK3CB,PIK3R1,PTPN ND signaling 11,SOS1,SPTAN1 Signaling by AKAP9,BRAF,EGFR,FN1,KRAS,N R-HSA-5654736 2.97E-05 F1,PIK3CA,PIK3CB,PIK3R1,PTEN ND FGFR1 ,PTPN11,SOS1,SPTAN1 AKAP9,BRAF,EGFR,FN1,KRAS,N Signaling by R-HSA-5654741 2.97E-05 F1,PIK3CA,PIK3CB,PIK3R1,PTEN ND FGFR3 ,PTPN11,SOS1,SPTAN1
35 Supplementary information, Wadi et al.
FDR of React.ID Description FDR Common genes Reactome.ID in 2010 AKAP9,BRAF,EGFR,FN1,KRAS,N Signaling by R-HSA-5654743 2.97E-05 F1,PIK3CA,PIK3CB,PIK3R1,PTEN ND FGFR4 ,PTPN11,SOS1,SPTAN1 Signaling by AKAP9,BRAF,EGFR,FN1,KRAS,N R-HSA-1433557 2.97E-05 F1,PIK3CA,PIK3CB,PIK3R1,PTEN ND SCF-KIT ,PTPN11,SOS1,SPTAN1 VEGFA-VEGFR2 AKAP9,BRAF,EGFR,FN1,KDR,KR R-HSA-4420097 2.97E-05 AS,NCF2,NF1,PIK3CA,PIK3CB,PI ND Pathway K3R1,SOS1,SPTAN1 AKAP9,BRAF,EGFR,FN1,KRAS,N Signaling by R-HSA-190236 3.07E-05 F1,PIK3CA,PIK3CB,PIK3R1,PTEN ND FGFR ,PTPN11,SOS1,SPTAN1 Signaling by AKAP9,BRAF,EGFR,FN1,KRAS,N R-HSA-5654738 3.07E-05 F1,PIK3CA,PIK3CB,PIK3R1,PTEN ND FGFR2 ,PTPN11,SOS1,SPTAN1 Signaling by AKAP9,BRAF,EGFR,FN1,KRAS,N R-HSA-1227986 3.18E-05 F1,PIK3CA,PIK3CB,PIK3R1,PTEN ND ERBB2 ,PTPN11,SOS1,SPTAN1 AKAP9,BRAF,EGFR,FN1,KDR,KR Signaling by R-HSA-194138 3.18E-05 AS,NCF2,NF1,PIK3CA,PIK3CB,PI ND VEGF K3R1,SOS1,SPTAN1 Downstream AKAP9,BRAF,EGFR,FN1,KRAS,N R-HSA-186763 signal 3.30E-05 F1,PIK3CA,PIK3CB,PIK3R1,PTEN 0.000443176 transduc on ,PTPN11,SOS1,SPTAN1
36