Linear models for microarray analysis
A gene expression experiment measures messenger RNA concentrations under a specific growth condition. We would like to know how does a gene concentration change when conditions change. But our measurements mix real gene expression and noise. How can we find separate noise and signal?
Linear models with categorical factors
Linear models allow us to predict from experimental data, and define confidence intervals for these predictions. Moreover, the coefficients of the linear model reveal useful information, with their corresponding confidence intervals. In this article we explore the case when some of the independent variables are not numeric but instead are in a nominal scale.
An easy way to transform your R script into a nice report
Let’s say we have written an R script and we want to include some comments about it. For example, we want to discuss the meaning of the results. It is easy to make a report from your R script, with the simple changes we discuss here.
Application: Uncertainty in Linear models
Beyond making predictions, linear models allow us to measure values that are hidden under a mountain of data. Learn how to measure these values and determine their confidence intervals.
Uncertainty in Linear Models
When we can assume that real values follow a straight line, we can reduce the uncertainty. Thus, we can predict the number of COVID-19 cases, and see if health-care decisions are effective or not.
How to solve hard problems
People are asking “How can I solve the homework? It is too hard”. Here are some steps that may help you.
Application: Evaluating Statistical Uncertainty
Measuring the same value several times may give different results. The real value is somewhere in a confidence interval. To find such interval we need to evaluate the statistical uncertainty.
Estimating Statistical Uncertainty
Every time we measure, we get a different number. How can we extract meaningful information from noisy data? How much information can we extract? Classical statistics gives an answer to these questions.
Some students argue that the exercise “find the largest value” is too hard. This raises the question “What do we mean by hard problem”? The answer teach us important things about life, the universe, and everything.
Comments on Homework 5
Several students have delivered their homework. Some are good, some are worth looking at them. I want to share them with everybody, so everybody can learn.
Three weeks of no-holiday
Universities in Turkey are closed for at least three weeks, as a way to contain the spread of Covid-19. It is easy to fall into the trap of doing “nothing”, watching TV, and read social media. But easy is not wise. Use your time wisely.
Methodology of Scientific Research
What is the value of Science? In this course we will discuss why do we need Science, how Science improves people’s lives and how to do it better.
Computing in Molecular Biology 2
This course is an introduction to Computational Thinking. We will use the tools we learned in the previous course and apply them to model and simulate scientific experiments as a way to understand them.
Should I learn Turkish? What is best for the University?
Some people ask me why I do not learn Turkish. It is a good question. For sure, it will be good for me. I would be able to speak about football with the butcher, discuss politics with my father-in-law, and read my collection of Matematik Dünyası magazine. But, what is best for our University?
My academic ancestors
A database of all mathematicians and their advisors shows my academic ancestors since Persia in 1200. Some famous names among my ancestors.
The Biologist Toolbox: Simulating Systems on the computer
Once we have a description of a system, and a nice drawing to represent it, we can answer some interesting questions. The first question is usually what will happen? In other words, we usually want to know what is the behavior of the system?.
The Biologist Toolbox: Drawing Systems
Can we represent a complex system in a simple and easy to understand way that is still realistic enough to be useful? The graphical representation that we discuss here, based Petri nets, can be used to summarize a dynamical system and can easily translated into a simulation.
Free fall experiment
We can use the computer to extract specific information from pictures and movies. Here we use some command-line tools to analyze a simple experiment. The experiment is simple, but the analysis had many unexpected compications.
Keep your mind sharp on August
Why do rich kids learn more than poor kids? According to some research, the “achievement gap” between kids of high-income and low-income families is not that rich kids learn more, but that they forget less in the summer holiday. Following this idea, I propose a couple of challenges to keep your mind active during the holidays, and maybe win a nice book.
My Professor got an International Prize
One of my professors got an international award. He had already won the most prestigious national recognitions in Science. Now he became global. This gives me the opportunity to start a series of posts about my world-class professors and their international achievements.