Limits...
A pan-cancer proteomic perspective on The Cancer Genome Atlas.

Akbani R, Ng PK, Werner HM, Shahmoradgoli M, Zhang F, Ju Z, Liu W, Yang JY, Yoshihara K, Li J, Ling S, Seviour EG, Ram PT, Minna JD, Diao L, Tong P, Heymach JV, Hill SM, Dondelinger F, Städler N, Byers LA, Meric-Bernstam F, Weinstein JN, Broom BM, Verhaak RG, Liang H, Mukherjee S, Lu Y, Mills GB - Nat Commun (2014)

Bottom Line: Therefore, direct study of the functional proteome has the potential to provide a wealth of information that complements and extends genomic, epigenomic and transcriptomic analysis in The Cancer Genome Atlas (TCGA) projects.The resultant proteomic data are integrated with genomic and transcriptomic analyses of the same samples to identify commonalities, differences, emergent pathways and network biology within and across tumour lineages.In addition, tissue-specific signals are reduced computationally to enhance biomarker and target discovery spanning multiple tumour lineages.

View Article: PubMed Central - PubMed

Affiliation: 1] Department of Bioinformatics and Computational Biology, 1400 Pressler St., The University of Texas MD Anderson Cancer Center, Houston, Texas 77030, USA [2].

ABSTRACT
Protein levels and function are poorly predicted by genomic and transcriptomic analysis of patient tumours. Therefore, direct study of the functional proteome has the potential to provide a wealth of information that complements and extends genomic, epigenomic and transcriptomic analysis in The Cancer Genome Atlas (TCGA) projects. Here we use reverse-phase protein arrays to analyse 3,467 patient samples from 11 TCGA 'Pan-Cancer' diseases, using 181 high-quality antibodies that target 128 total proteins and 53 post-translationally modified proteins. The resultant proteomic data are integrated with genomic and transcriptomic analyses of the same samples to identify commonalities, differences, emergent pathways and network biology within and across tumour lineages. In addition, tissue-specific signals are reduced computationally to enhance biomarker and target discovery spanning multiple tumour lineages. This integrative analysis, with an emphasis on pathways and potentially actionable proteins, provides a framework for determining the prognostic, predictive and therapeutic relevance of the functional proteome.

Show MeSH

Related in: MedlinePlus

Unbiased data-driven signaling networkUnbiased signaling network based on a probabilistic graphical models analysis, visualizing all 11 tumor lineages individually. Interplay between nodes was quantified using scores from the graphical model analysis (see Methods), that identify links between nodes whilst controlling for the effects of all other observed nodes. The strength of links is indicated by the thickness of the line whilst the color indicates the tumor lineage in which the link was observed; only the strongest links are shown. Nodes in white are related nodes that were highly correlated and therefore merged prior to network analysis. The adjacent correlated (green) node was then used for network generation. Positive (negative) correlations are indicated with continuous (dotted) lines. A high-resolution image of the network can be found online (http://bioinformatics.mdanderson.org/main/TCGA/Pancan11/RPPA).
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4109726&req=5

Figure 6: Unbiased data-driven signaling networkUnbiased signaling network based on a probabilistic graphical models analysis, visualizing all 11 tumor lineages individually. Interplay between nodes was quantified using scores from the graphical model analysis (see Methods), that identify links between nodes whilst controlling for the effects of all other observed nodes. The strength of links is indicated by the thickness of the line whilst the color indicates the tumor lineage in which the link was observed; only the strongest links are shown. Nodes in white are related nodes that were highly correlated and therefore merged prior to network analysis. The adjacent correlated (green) node was then used for network generation. Positive (negative) correlations are indicated with continuous (dotted) lines. A high-resolution image of the network can be found online (http://bioinformatics.mdanderson.org/main/TCGA/Pancan11/RPPA).

Mentions: Based on the availability of protein data across a large number of samples, we used a probabilistic graphical models approach44, 45 without the inclusion of prior knowledge to create an unbiased signaling network (Fig. 6, see Methods). We used the relatively large number of samples per tumor lineage to elucidate links in specific cancers and across multiple cancers, inferring networks using tumor lineage-specific samples. Interplay between nodes was quantified using scores from the graphical model analysis (see Methods) that identify links between nodes whilst controlling for the effects of all other observed nodes. Several expected links were observed across most tumor types, including pMEK with pERK, beta-catenin with E-cadherin and pPKCdelta with pPKCalpha and pPKCbeta, supporting the ability of RPPA analysis to yield high-quality signaling information from TCGA samples. Other expected links were seen in only a subset of tumors such as pAKT with pPRAS40 and pTSC2 (tuberinPT1462), consistent with differential wiring of signaling pathways in different cancers. A number of other links such as MYH11 with Rictor, cyclinB1 with FOXM1, and pACC with FASN were not expected and warrant further exploration. The interplay between p85 and PTEN is consistent with our demonstration that p85 is a key determinant of PTEN stability39, 46. The negative link between pAKT and PTEN was expected, but the one between p85 and claudin7 in LUSC was not and may be worthy of further exploration. PI3K/AKT signaling does not link clearly to mTOR, which appears to primarily be downstream of MAPK signaling47, 48, 49. The relatively weak links in the PIK3K/AKT pathway are striking given the degree of antibody representation for this pathway in the RPPA analysis. Key nodes such as CDK1 unexpectedly linked a wide range of protein pathways. Overall, the data suggest that the EGFR receptor family, together with the linked MEK and MAPK pathways, is the dominant determinant of signaling across the cancer lineages in the Pan-Cancer analysis. Using independent datasets in breast cancer, ovarian cancer and endometrial cancer, as well as published research, many of the strongest protein links in the network could be validated (Supplementary Fig. 13 and Supplementary Table 15), supporting the notion that large RPPA-based protein datasets can be used to “learn” networks in an unbiased manner.


A pan-cancer proteomic perspective on The Cancer Genome Atlas.

Akbani R, Ng PK, Werner HM, Shahmoradgoli M, Zhang F, Ju Z, Liu W, Yang JY, Yoshihara K, Li J, Ling S, Seviour EG, Ram PT, Minna JD, Diao L, Tong P, Heymach JV, Hill SM, Dondelinger F, Städler N, Byers LA, Meric-Bernstam F, Weinstein JN, Broom BM, Verhaak RG, Liang H, Mukherjee S, Lu Y, Mills GB - Nat Commun (2014)

Unbiased data-driven signaling networkUnbiased signaling network based on a probabilistic graphical models analysis, visualizing all 11 tumor lineages individually. Interplay between nodes was quantified using scores from the graphical model analysis (see Methods), that identify links between nodes whilst controlling for the effects of all other observed nodes. The strength of links is indicated by the thickness of the line whilst the color indicates the tumor lineage in which the link was observed; only the strongest links are shown. Nodes in white are related nodes that were highly correlated and therefore merged prior to network analysis. The adjacent correlated (green) node was then used for network generation. Positive (negative) correlations are indicated with continuous (dotted) lines. A high-resolution image of the network can be found online (http://bioinformatics.mdanderson.org/main/TCGA/Pancan11/RPPA).
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4109726&req=5

Figure 6: Unbiased data-driven signaling networkUnbiased signaling network based on a probabilistic graphical models analysis, visualizing all 11 tumor lineages individually. Interplay between nodes was quantified using scores from the graphical model analysis (see Methods), that identify links between nodes whilst controlling for the effects of all other observed nodes. The strength of links is indicated by the thickness of the line whilst the color indicates the tumor lineage in which the link was observed; only the strongest links are shown. Nodes in white are related nodes that were highly correlated and therefore merged prior to network analysis. The adjacent correlated (green) node was then used for network generation. Positive (negative) correlations are indicated with continuous (dotted) lines. A high-resolution image of the network can be found online (http://bioinformatics.mdanderson.org/main/TCGA/Pancan11/RPPA).
Mentions: Based on the availability of protein data across a large number of samples, we used a probabilistic graphical models approach44, 45 without the inclusion of prior knowledge to create an unbiased signaling network (Fig. 6, see Methods). We used the relatively large number of samples per tumor lineage to elucidate links in specific cancers and across multiple cancers, inferring networks using tumor lineage-specific samples. Interplay between nodes was quantified using scores from the graphical model analysis (see Methods) that identify links between nodes whilst controlling for the effects of all other observed nodes. Several expected links were observed across most tumor types, including pMEK with pERK, beta-catenin with E-cadherin and pPKCdelta with pPKCalpha and pPKCbeta, supporting the ability of RPPA analysis to yield high-quality signaling information from TCGA samples. Other expected links were seen in only a subset of tumors such as pAKT with pPRAS40 and pTSC2 (tuberinPT1462), consistent with differential wiring of signaling pathways in different cancers. A number of other links such as MYH11 with Rictor, cyclinB1 with FOXM1, and pACC with FASN were not expected and warrant further exploration. The interplay between p85 and PTEN is consistent with our demonstration that p85 is a key determinant of PTEN stability39, 46. The negative link between pAKT and PTEN was expected, but the one between p85 and claudin7 in LUSC was not and may be worthy of further exploration. PI3K/AKT signaling does not link clearly to mTOR, which appears to primarily be downstream of MAPK signaling47, 48, 49. The relatively weak links in the PIK3K/AKT pathway are striking given the degree of antibody representation for this pathway in the RPPA analysis. Key nodes such as CDK1 unexpectedly linked a wide range of protein pathways. Overall, the data suggest that the EGFR receptor family, together with the linked MEK and MAPK pathways, is the dominant determinant of signaling across the cancer lineages in the Pan-Cancer analysis. Using independent datasets in breast cancer, ovarian cancer and endometrial cancer, as well as published research, many of the strongest protein links in the network could be validated (Supplementary Fig. 13 and Supplementary Table 15), supporting the notion that large RPPA-based protein datasets can be used to “learn” networks in an unbiased manner.

Bottom Line: Therefore, direct study of the functional proteome has the potential to provide a wealth of information that complements and extends genomic, epigenomic and transcriptomic analysis in The Cancer Genome Atlas (TCGA) projects.The resultant proteomic data are integrated with genomic and transcriptomic analyses of the same samples to identify commonalities, differences, emergent pathways and network biology within and across tumour lineages.In addition, tissue-specific signals are reduced computationally to enhance biomarker and target discovery spanning multiple tumour lineages.

View Article: PubMed Central - PubMed

Affiliation: 1] Department of Bioinformatics and Computational Biology, 1400 Pressler St., The University of Texas MD Anderson Cancer Center, Houston, Texas 77030, USA [2].

ABSTRACT
Protein levels and function are poorly predicted by genomic and transcriptomic analysis of patient tumours. Therefore, direct study of the functional proteome has the potential to provide a wealth of information that complements and extends genomic, epigenomic and transcriptomic analysis in The Cancer Genome Atlas (TCGA) projects. Here we use reverse-phase protein arrays to analyse 3,467 patient samples from 11 TCGA 'Pan-Cancer' diseases, using 181 high-quality antibodies that target 128 total proteins and 53 post-translationally modified proteins. The resultant proteomic data are integrated with genomic and transcriptomic analyses of the same samples to identify commonalities, differences, emergent pathways and network biology within and across tumour lineages. In addition, tissue-specific signals are reduced computationally to enhance biomarker and target discovery spanning multiple tumour lineages. This integrative analysis, with an emphasis on pathways and potentially actionable proteins, provides a framework for determining the prognostic, predictive and therapeutic relevance of the functional proteome.

Show MeSH
Related in: MedlinePlus