Proteins and Human Disease

Proteins and Human Disease: Insights and Applications

• Understanding the pathways that impact phenotype is crucial after identifying a specific gene linked to disease.
• Diseases such as cystic fibrosis (CFTR), sickle-cell anemia (hemoglobin beta), "mad cow" disease (prion protein), and Alzheimer's disease (amyloid precursor protein) are influenced by proteins.
• Single amino acid substitutions can dramatically alter protein structure, such as the DF508 mutation in CFTR and the E6V mutation in hemoglobin beta.
• Modeling protein structure, function, and interactions is essential for understanding disease phenotypes, risk, and drug targets.
• Websites like ExPASy and ISREC provide resources for analyzing protein properties, but their accuracy varies.
• Protein secondary structure is determined by amino acid side chains, with examples including a-helices in myoglobin and b-sheets in thioredoxin.
• Prediction algorithms like Chou and Fasman's and modern methods use multiple sequence alignments and machine learning techniques to predict secondary protein structure.
• Approaches for determining tertiary protein structure include experimental methods (X-ray crystallography, NMR) and prediction methods like comparative modeling and ab initio prediction.
• The Protein Data Bank (PDB) is a key repository for protein structures, containing 97,000 entities in 2014.
• The CATH Hierarchy of Structure and tools like VAST and JMol provide access to structure data and comparisons of protein structures.
• More complex models have been developed to model molecular interactions, including protein-protein and protein-DNA interactions.
• Drug discovery involves identifying compounds that can alter protein interactions, with steps including finding genomic regions associated with disease and determining the effect on protein structure and function.

Gene Expression Analysis: Microarrays and RNAseq

• Gene expression analysis involves comparing RNA samples from different conditions, such as normal colon and colon tumor.
• Two main methods for gene expression analysis are microarrays and RNAseq, with the latter being genome-wide and measuring all RNA transcripts.
• Microarrays involve short oligonucleotide probes representing exons and fluorescence-labeled cDNA hybridization, followed by image acquisition and data analysis.
• The seven main steps for differential gene expression on microarrays include experimental design, RNA extraction, hybridization, image analysis, data analysis, biological confirmation, and depositing in databases.
• Key considerations in microarray analysis include the importance of biological replicates, RNA and cDNA quality, and the impact of technical artifacts on results.
• Microarray data analysis involves statistical analysis to determine differentially expressed RNA transcripts, as well as clustering and classification to identify meaningful biological patterns.
• Biological confirmation of microarray results is crucial, typically using methods such as Northern blots, Western blots, RT-PCR, or RNAseq.
• Two main repositories for microarray data are the Gene Expression Omnibus (GEO) at NCBI and ArrayExpress at the European Bioinformatics Institute, with specific requirements for experimental information (MIAME).
• GEO Datasets allow users to search and analyze RNA transcript expression across experiments, perform statistical tests, and create heatmaps and gene expression clusters.
• RNAseq involves isolating RNAs, converting to cDNA, sequencing, mapping to the genome, and downstream analysis of differential gene expression.
• RNA studies require complex design due to technical and biological replicates, and the need to account for variation within cells, individuals, and environmental factors.
• RNAseq pipeline includes steps like read alignment, transcript compilation, gene identification, and differential expression, with the goal of achieving high correlation coefficients between replicates.