Molecular Pathway Analysis Using High-Throughput OMICS Molecular Data

Anton Buzdin is professor and head of laboratories at Sechenov First Moscow Medical University, Moscow Institute of Physics and Technology, and Shemyakin-Ovchinnikov Institute of Bioorganic Chemistry. He was elected as the chair of Biostatistics and Bioinformatics subgroup, Pathobiology Group of European Organization for Research and Treatment of Cancer (EORTC). Professor Buzdin is cofounder and chief scientific officer of Oncobox, which is a translational oncology company. He has authored more than 170 scientific papers and patent applications, and has worked as chief executive of several biotechnological and biomedical companies in Russia, Hong Kong and in the United States. He has also chaired two clinical trials of transcriptomics-based diagnostic tools in oncology. Professor Buzdin’s team pioneered bioinformatic quantitative analysis of molecular pathways using gene expression data and developed specialized algorithms and software packages to personalize prescription of targeted immunotherapy and chemotherapy drugs to cancer patients that were clinically validated and commercialized.

Contributors
Preface

PART I: Foundational information

Chapter 1: Past, current, and future of molecular pathway analysis
Anton Buzdin, Alexander Modestov, Daniil Luppov and Ira-Ida Skvortsova

1.1. Molecular pathways
1.2. Quantitative omics data
1.3. Different levels of omics data analysis
1.4. Quantization of IMP activities
1.4.1. Annotation of functional roles for pathway participants
1.5. Applications of IMP analysis
1.5.1. Applications in medicine
1.6. Software for quantitative assessment of IMP activation
1.7. Concluding remarks
References

Chapter 2: Molecular data for the pathway analysis
Xinmin Li and Anton Buzdin

2.1. Omics data available for the molecular pathway analysis
2.2. Data needed to reconstruct IMPs
2.3. Data needed to estimate activation levels of IMPs
References

Chapter 3: Benefits and challenges of OMICS data integration at the pathway level
Nicolas Borisov and Maksim Sorokin

3.1. Background
3.2. The comparison
3.2.1. Functional annotation of gene expression data
3.2.2. Statistical tests
3.2.3. Mathematical modeling
3.2.4. Analysis of gene expression datasets
3.2.5. Biological relevance of cross-platform harmonized expression data
3.2.6. Marker gene and pathway analysis
3.3. Results
3.3.1. Cross-platform processing of transcriptomic and proteomic data
3.3.2. Building pathway activation profiles and assessment of batch effects
3.3.3. Mathematical modeling of data aggregation effects
3.3.4. Experimental model of cross-platform comparisons
3.3.5. Data aggregation effects assessed for RNA and protein expression levels
3.3.6. Comparison of data aggregation capacities of different PAL scoring methods
3.3.7. Retention of biological features
3.3.8. Gene and pathway analysis of PTSD datasets
3.4. Discussion
Abbreviations
References

Chapter 4: Controls for the molecular data: Normalization, harmonization, and quality thresholds
Nicolas Borisov

4.1. Background
4.2. Principles of harmonization algorithms
4.3. Differential clustering of human normal and cancer expression profiles
4.4. Correlation, regression, and sign-change analysis of cancer drug balanced efficiency score (BES) after application of different methods of harmonization
4.5. Discussion
Abbreviations
References

Chapter 5: Reconstruction of molecular pathways
Anton Buzdin and Maksim Sorokin

5.1. Molecular pathways
5.2. An approach to reconstruct the pathway
5.2.1. The interactome model
5.2.2. Building gene-centric pathways
5.2.3. Overall functional annotation of reconstructed pathwaysdgene ontology classification
5.2.4. Visual annotation of reconstructed pathways
5.2.5. Algorithmic annotation of functional roles for pathway components
5.2.6. Examples of building and annotation of molecular pathways
References

Chapter 6: Qualitative and quantitative molecular pathway analysis: Mathematical methods and algorithms
Nicolas Borisov, Stella Liberman-Aronov, Igor Kovalchuk and Anton Buzdin

6.1. Background
6.2. Topology-based methods for pathway activation assessment
6.2.1. Oncobox
6.2.2. Topology analysis of pathway phenotype association
6.2.3. Topology-based score
6.2.4. Pathway-express
6.2.5. Signal pathway impact analysis
6.2.6. iPANDA (in silico pathway activation network decomposition analysis)
6.3. Methods for database preparation for pathway activation assessment
6.3.1. Curation of pathway databases
6.3.2. Algorithmic annotation of pathway graph nodes
6.3.3. Finding gene importance factors for iPANDA
6.4. Personalized ranking of cancer drugs based on PALs
6.4.1. Oncobox balance efficiency score (BES)
6.4.2. Drug efficiency index (DEI)
6.5. Multi-omics data pathway analysis
6.5.1. Pathway activation assessment for methylome, microRNAs, and long noncoding (LNC) antisense (AS) RNAs
6.6. Concluding remarks
Abbreviations
References
Further reading

PART II: Methods and guidelines

Chapter 7: Getting started with the molecular pathway analysis
Anton Buzdin and Xinmin Li

7.1. Strategies of pathway analysis
7.2. Reconstruction of pathways and networks
7.3. The devil is in the things
7.4. Applications of molecular pathway analysis
7.5. Preprocessing of data for pathway analysis
7.6. Visualization of pathways
References

Chapter 8: Molecular pathway analysis using comparative genomic and epigenomic data
Ye Wang, Marianna Zolotovskaia and Anton Buzdin

8.1. Types of pathway analysis requiring (epi)genomic data
8.2. Profiling of genomic pathway instability by using DNA mutation data
8.2.1. Initial mutation data
8.2.2. Algorithm validation dataset
8.2.3. Molecular target interrogation dataset
8.2.4. Clinical trial data
8.2.5. Molecular pathway data
8.2.6. Pathway instability scoring
8.2.7. PI analysis of cancer mutation signatures
8.2.8. PI-based drug scoring
8.2.9. Assessment of MDS family methods performance using clinical trial data
8.2.10. Application of MDS to identify putative drug target genes
8.3. Epigenetic marks as the measure of IMP molecular evolution
8.3.1. Study design
8.3.2. Source IMPs
8.3.3. Aggregated dN/dS data
8.3.4. RE regulation enrichment data
8.3.5. Functional classification of histone modifications
8.3.6. Aggregated NGRE score
8.3.7. Correlation between structural and regulatory evolutionary rate metrics
8.3.8. Functional groups of genes and pathways with different evolutionary rates
8.4. Concluding remarks
References

Chapter 9: Quantitative molecular pathway analysis using transcriptomic and proteomic data
Anton Buzdin, Sergey Moshkovskii and Maksim Sorokin

9.1. Types of molecular pathway analysis
9.2. Quantitative analysis of gene expression
9.3. Quantitative assessment of the pathway activities
9.3.1. Calculation of PAL
9.3.2. Annotation of functional roles of IMP members
9.4. Software
9.4.1. Visualization of the pathways
9.4.2. Manual on the installation of oncoboxlib library
References

Chapter 10: MicroRNA data for quantitative analysis of molecular pathways
Anton Buzdin and Alina Artcibasova

10.1. Relevance of microRNA profiles to molecular pathway activation analysis
10.2. Algorithmic analysis of pathway activation
10.3. Applications of pathway analysis for microRNAs
10.3.1. MiRImpact application to profile regulation of IMPs in bladder cancer
10.3.2. MiRImpact application to profile regulation of IMPs during cytomegaloviral infection
10.4. Concluding remarks
References

Chapter 11: Methods and tools for OMICS data integration
Ilya Belalov and Xinmin Li

11.1. A snapshot of the current state of OMICS integration landscape
11.2. The most important part of this chapter
11.3. Best practices in preprocessing multiomics datasets
11.4. OMICS data integration in the eyes of a life scientist
11.4.1. From genotype to phenotype: Step I-Transcription
11.4.2. From genotype to phenotype: Step II-Translation
11.4.3. From genotype to phenotype: Step III-Proteins
11.4.4. From genotype to phenotype: Step IV-Metabolites
11.5. Data scientist summary
11.6. Life scientist summary
References
Further reading

PART III: Practical applications

Chapter 12: Molecular pathway approach in clinical oncology
Anton Buzdin, Alexander Seryakov, Marianna Zolotovskaia, Maksim Sorokin, Victor Tkachev and Alf Giese

12.1. Gene expression data in clinical oncology
12.2. Conversion of pathway activation data into personalized prediction of cancer drug efficacy
12.2.1. Molecular pathway databank
12.2.2. Clinical trial database
12.2.3. Drug target database
12.2.4. Algorithmic scoring of cancer drug efficiencies
12.3. Examples of IMP-based clinical ranking of drugs in oncology
12.3.1. Example 1. Ranking of cancer drugs based on mRNA expression data
12.3.2. Example 2. Comparison of alternative drug scoring methods
12.4. Conclusion
References

Chapter 13: Molecular pathway approach in pharmaceutics
Anton Buzdin, Teresa Steinbichler and Maksim Sorokin

13.1. Molecular pathway analysis in general
13.1.1. What is intracellular molecular pathway
13.1.2. Molecular pathway analysis
13.1.3. Pathway analysis instruments
13.2. Pathway analysis to facilitate tasks in molecular pharmacology
13.2.1. Task 1. To establish mechanism of action of drug candidate X
13.2.2. Task 2. To identify robust response biomarkers for drug (candidate) X
13.2.3. Task 3. To identify drugs that act similarly to drug (candidate) X or to identify molecular targets of X
13.3. Practical examples how IMP analysis may help
13.4. Useful online resources
13.5. Conclusion
References

Chapter 14: Molecular pathway approach in biotechnology
Anton Buzdin, Denis Kuzmin and Ivana Jovcevska

14.1. Pathways of biotechnology
14.1.1. Biotechnology
14.1.2. The pathways
14.1.3. Molecular pathways in biotech
14.2. Examples of pathway analysis in biotechnology
14.2.1. Golden rice
14.2.2. A humanized N-glycosylation system for expression of human proteins in yeast
14.2.3. Optimization of the photosynthesis system
14.3. Conclusion and perspective
References

Chapter 15: Molecular pathway approach in biology and fundamental medicine
Anton Buzdin, Ye Wang, Ivana Jovcevska and Betul Karademir-Yilmaz

15.1. Molecular pathway analysis in biomedicine
15.2. IMP analysis in oncology
15.2.1. IMPs in cancer
15.2.2. Quantitative analysis of IMPs in oncology
15.2.3. IMPs as cancer biomarkers
15.2.4. Pathway-based scoring of cancer drug efficiencies
15.3. Other applications of IMP analysis in biomedicine
15.3.1. Ranking and repurposing of drugs
15.3.2. Understanding molecular mechanisms
15.4. Conclusion
References