Preface

This is an evolving, never-complete notebook for biological marker (biomarker) analysis in the biomedical setting. The purpose is to document the concept of each type of biomarker and analytical approach. In general, it covers the topics of DNA, RNA, protein, and cell types analyses. It also talks about some statistical issues such as multiplicity, power calculation, detection limit, etc that one may face when analyzing biomarker data.

Additionally, I provide R code examples for implementing methods. These codes are basic and are for illustration purposes only. This book is not meant for teaching purposes, albeit scientific references and resources are provided in each chapter. The analytical approaches in this book are based on my personal work experience and are not subject to any kind of convention in any enterprise setting. It is open to constructive suggestions. I do not have a strong urge to recommend any analytical approach in real practice since each problem is unique. However, the fundamental principles are discussed in the context of these notes. It is up to the analytical personnel to decide based on the situations they are facing.

Last but not least,

Markdown, no MS Word. ☀️

—- awesome tomato

Author

Became a data scientist

Towards the end of my second postdoctoral work, I decided to leave academia. Almost a decade of training in research left me nothing so handy for any industrial professions but an analytical mind and skills for learning. Becoming a data scientist was not a coincidence. I like statistics as well as methodologies and I like to deal with data. I didn’t know what data science was, but I was told that I have a data scientist profile. So I took my first non-academic job as a data scientist.

How did I started programing?

I was, am, will be ‘lazy’. I don’t like to repeat things, especially something like redoing analysis by just changing a predictor, or producing the same table with n different ways. So, while cranking and juggling among three research projects and ‘dealing with’ my advisors’ ‘what ifs…’, I figured out a way to do less and produce more by compressing these repetitions into functions. I even created functions directly output result tables so that I didn’t need to spend hours perfecting them on MS word. I didn’t tell anyone about this because I was afraid that if they knew they would ask me to do more analyses. Also, I didn’t think it’s a big deal, until one day my friend/colleague said to me, I wish I could have your programming skills. Man, it sounded so cool, even though I was just writing a few comment lines on the screen … 😎 Then as computations got heavier, I learned parallel computing, logging, etc. I started to enjoy let the computer/server do the work and I was physically somewhere else. Eventually, I was able to build many things according to my needs. It’s like building a robot and having control over it. I like the feeling of controlling something, I don’t even get to control my life most of the time. But I’m a full commander of the programs. (Well, when it’s not bugging.)

More background

You probably will understand me better by reading my blogs. Just in case this is not enough to make a professional decision, Here is my CV website. Yes, a website just for my curriculum vitae! Spoiler alert, it’s boring.