In-silico Medicine & Biostatistics

• Medical informatics and biostatistics are key areas of study
• Andrzej Bąk can be contacted via [email protected] or [email protected], or by phone at 32 359 11 97
• The office is located at Szkolna 9 Street, Katowice, office number 40, floor 1

In-silico Medicine Outline

• Encompasses data, informatics, and knowledge in medicine
• Molecular editors code drug structures
• Receptors serve as molecular targets
• Molecular descriptors indicate lipophilicity
• CAMD (Computer Assisted Molecular Design) facilitates QSAR modeling
• AI (Artificial Intelligence) in medicine includes neural network essentials
• Includes basic biostatistics, data preprocessing, data modeling, and techniques of data visualization and grouping

Computer Assisted Drug Design (CADD)

• Aims to produce substances characterized by a desired action profile in the human body
• Leads to rational production of medicines
• Enables the "molecular production of properties" through drug modeling

Defining "Medicine"

• It's the art of human treatment or healing
• It includes pharmacy, pharmacology, and medicinal chemistry
• All these sciences deals with molecules and their transformations
• The key question is, what constitutes a drug molecule?
• "All substances are poisons; it is the dose that differentiates a poison from a remedy" - Paracelsus, 16th century

Drug Discovery History

• Around 5100 years ago in China, Chiang Shan herbs (for fever) and Ma Huang (anti-cough) were identified
• 3 w.p.n.e saw extraction of pain-killers from poppy grain
• Ancient Greece yielded the discovery of scopolamine (truth serum) from Belladonna
• Inca nations used coca for energizing purposes
• India discovered rezorcinol used to treat blood pressure
• In 1633, a Calancha monk extracted a drug to treat malaria from quina tree
• In 1928, Aleksander Fleming discovered penicillin (antibiotics)
• Fleming stated that "The Nature produced penicilline, I have just discoverd it"
• It is often possible to unearth something not sought

Serendipitous Discoveries

• Sometimes are the result of a "lucky shot"
• Study of biological discharge led to the discovery that it could kill bacteria
• Lisozyme became the first antibiotic in 1921
• Penicillin was discovered in 1928 from Penicillium notatum fungus, which affects staphylococcus
• In 1941, Florey and Chain experimented on mice with penicillin
• In 1941, Albert Aleksander healed a patient with a scratch on his face made by a thorn, but ultimately the patient died
• Afterward, Karl Folkers, working at Merk, specified the structure of penicyline

Medicinal Product Definition

• According to Directive 2001/83/EC, it's any substance or mixture to treat or prevent disease in humans or animals modifying physiological functions
• Colloquially, medicine is a substance that modifies physiological processes
• It can inhibit disease causes or symptoms, prevent disease development, or aid in diagnostic purposes
• Medicines can also modify unchanged functions, e.g., contraceptive drugs

Pharmacy and Pharmacology

• Pharmacy is the science of drug production and interactions
• Pharmacology studies the therapeutic application, mechanisms, and side effects of drugs
• In Poland pharmaceutic law regulates drug rules
• Polish law doesn't refer to "drugs", but to "healing/treating/medical products"
• The Department of Drug Registration, Medical Products and Biokilling Products grants permission for drug sales, after approval by the Health Department head
• Drug types include innovative and generic

Drug Molecule Classification

• Remedial agents are categorized by origin
• Natural compounds: from plants/animals like vitamins, hormones, amino acids, antibiotics, alkaloids, glycosides
• Synthesis compounds: pure or naturally occurring synthesis, like morphine, atropine, steroids, and cocaine, used to reduce costs
• Semi-synthesis compounds: derived when synthesis/isolation is costly
• Natural intermediates of these drugs are used to synthesize desired products like semi-synthetic penicillins

Data Exploration: Expectations vs. Reality

• Expectations include discovering new leading structures and many new drugs
• Reality often involves experimental issues, few new drugs, and high production costs

High Drug Costs

• The expense stems from absorption, distribution, metabolism, excretion, and toxicity (ADMETox) studies
• Pharmaceutical phase: Molecules disintegrate as drug availability depends on drug form and administration method
• Pharmacokinetic phase: Drug transport occurs through biological barriers affecting availability
• Pharmakodynamic phase: Interactions with macromolecular targets (drug-receptor complex creation)

Drug Costs & HIV Struggle

• Generic drugs aim to undercut branded drugs from pharmaceuticals like Pfizer & Lilly
• Africa has 30 million HIV infections a year
• HAART therapy costs $15,000 while generic versions cost < $1/day
• Company patent rules impact drug availability
• 70% of drugs are discovered by academia
• 1/3 of funds go to research, while the rest is spent on ads and salaries
• In the USA, patent law became more restricted despite gains in the HIV battle
• In 2000, 8,000 people in Africa were treated, while in 2010, over 1 million were treated
• Each year, 17 million people die from treatable disease
• Developing a new drug costs $1 billion and takes 10 years, plus 5 years of patent protection, as well as phases I-IV of clinical studies

Drug Discovery Cascade

• Begins with High-Throughput Screening (HTS) and progresses to drug candidates
• Failure risk goes up during the development

Data Analysis as Knowledge

• Employs experimental measures and computer-assisted simulations
• Involves 150 million molecules, about 1 million yearly, and 800,000 publications annually
• Structures are evaluated using numeric and spectral data
• Deductive learning relies on calculations
• Inductive learning relies on observation for modeling
• Data exploration refines data into information and knowledge

The Value of Knowledge Discovery

• Knowledge is stored, sent, and processed
• Information is the most valuable asset
• Students have to learn everything as part of society
• PhDs know where to find important information
• Professors know where to find doctors

The Need for New Molecules

• Objective of synthesis is not the production of new compounds but properties
• The principal aim of in-silico medicine is the molecular production of properties

Computer Science in In-Silico Medicine

• Involves informatics methods to solve drug design and interaction problems

Informatics in Medicine

• Involves the coding of drug structures and properties in PC programs
• Medical databases and data exploration procedures
• Calculations of various drug parameters
• Includes prediction of interactions using molecular docking
• Statistical analysis in medicine using biostatistics
• Also provides computer procedures and software

In-Silico Medicine Process

• It starts with defining a problem, building models, making calculations, and analyzing results

Drug Analysis Process

• Begins with 2D to 3D molecular generation, followed by pre-processing/descriptor selection, and then structure coding
• Utilizes statistical and artificial intelligence methods via neural networks for data analysis

Chemoinformatics of Drug Molecules

• Cycle including molecule, structure, descriptors, data, databases, measures and database exploration

Cheminformatics and Bioformatics

• Using pharmacophore, serendipity, and de novo pathways
• Molecular docking occurs
• Hit leads to drugs for a target
• Uses similarity searching

Two Approaches to Drug Design

• Ligand-based Drug Discovery (RI) and Structure-based Drug Discovery (RD)
• RI uses Ligands collection maps of interactions to reverse image of target structure to lead to selection, optymalization, and clinical studies
• RD searches for the molecular weight and complementary fullfilment for target molecule to undergo identyfication, preparation, docking ,and evaluation

CAMD Study Directions

• Involves preparation of receptor and ligand, and structural filters include drug likeness
• Includes docking studies and ligand-receptor geometry
• Encompasses post-processing and calculating activity prediciton
• Focuses on Pose evaluation and chemical stability

Drug Implementation Stages

• Specifying the disease and lead identification
• Improves with optimization
• Proceeds from interaction place determination towards identification of pharmacophore
• Uses selection of biological testing that requires subsequent lead identification for metabolic and toxic or clinical studies

Drug Design Evolution

• Goes from searching for synthetic analogues of natural products to leading structures
• This moves towards personal drug design through translational medicine

Virtual Drug Coding Challenges

• These include storing molecular structures, searching databases and treating chemical reactions, with synthesis planning, all to derive properties

Molecular Presentation

• Can be done using 3D models

Stereochemistry's Importance

• Thalidomide's (R)-enantiomer has sedative effects
• The (S)-enantiomer's an embryo-toxic and teratogenic agent
• Enantiomers can racemize in vivo

Linear Notation in Drug Coding

• Uses representations like SMILES and SLN
• Example given is the molecule Phenylalanine and it's representations along with it's IUPAC name

Molecular Surfaces and Drug Design

• Molecular surfaces can be represented in different methods
• Surfaces can be combined with the presentation

Graph Theory

• Euler's task in Königsberg was to find a walk through the town crossing each of the seven bridges exactly once and finishing at the starting point
• Drugs can be modeled using this approach

Drug Coding

• Uses neighboring matrix/bondig matrix, atom list, and connection table