Hello, and welcome to Biostat Squid. In this video, I will be explaining what volcano plots are and how to interpret them. So, let's get started.

Let's imagine you're carrying out an RNA experiment. You have a group of cells, A, and a group of cells, B. Group B was additionally treated with a drug.

Now, you want to see what effect the drug has on gene expression. Does the drug cause some genes to be upregulated? Does it downregulate the expression of other genes? By how much? A great way to visualize this is using a volcano plot.

A volcano plot is just a type of scatter plot that shows statistical significance (usually represented by p-values) versus magnitude of change (usually represented by the fold change).

So, in this case, each dot represents a gene. This way, we can quickly identify genes with large fold changes that are also statistically significant. These might be the most biologically significant genes of your experiment.

Now, how do we interpret a volcano plot? The y-axis shows the level of statistical significance. It is usually represented by the p-value, which means the most statistically significant genes are towards the top because they have the lowest p-values.

higher in Group A compared to Group B, then the fold change will be negative. So the most downregulated genes are towards the left; they have a negative fold change. This means that they have a lower expression in Group B compared to A. And of course, if the fold change nears

zero, then the expression does not really change between one group and another. Now, you might have noticed that the fold change is usually converted to the logarithm to base 2. This is because it makes things easier to interpret, and it's easier to represent a wider range

of values. So, let's give an example: if the expression of Gene 1 is twice as intense in Group B versus Group A, then the fold change is 2, meaning it's twice as big in B than in A. But the log2 fold

change will be 1. If the expression is four times as big, then the log2 fold change is 2. If the expression is eight times as big, then the log2 fold change is 3. Then the same happens when the change decreases: if the expression is half as big in

Group B compared to Group A, that is equal to a log2 fold change of -1. A quartering would be equal to a log2 fold change of -2, and so on. Let's talk a bit more about the p-values in the y-axis. As you might have noticed,

the negative log10 p-value is the most common way of representing p-values on a volcano plot. But why do we use negative log10 p-value? Well, of course you can also use raw p-values, but aesthetically it doesn't look that nice. Although, to be

Fair. They look a bit more like volcano plots. Um, this is because p-values range from 0 to 1, and highly significant genes, which are usually the genes you're interested in, they have really low p-values. We're talking about, um, lots and lots of zeros. For example, a

p-value of an interesting gene might be that big number, or small number I should say, that you see there. But if you see, but if you take the minus log10, we get 22. So it's a much nicer number to handle. So, remember, a volcano plot is used to

easily visualize which are the most significant differences and how big is that difference between two groups. The y-axis shows if that change is significant or not, and it is usually in the form of minus log10 p-value. The x-axis shows the fold change: how many

times are genes expressed more or less compared to the other group. If you want to check highly significant genes, they will have higher log10 p-values. If you want downregulated genes, you usually look to the left, to negative fold changes. Upregulated genes will then be to the

right, with positive fold changes. So this is the end of this video. I hope you liked it. If you found it useful, do let me know in the comments down below. Don't forget to like and subscribe, and if you would like more content, do let me know. Your feedback is really important

to me, and that is all for today. So see you in the next one.

Volcano plots explained | How to interpret a volcano plot for DGE

Welcome to Biostatsquid!
In this channel, you will find clear explanations and hands-on tutorials for analysing and interpreting biological data. 
- Learn basic techniques in biostatistics, from heatmaps to pathway enrichment analysis.
- Manage your data and create publication-ready visualisations by following my easy step-by-step tutorials in R and Python.
- Practical explanations without complex formulas for you to understand your datasets and be able to interpret your results.
Are you ready to dive in? Don't miss any other video and hit that subscribe button!


Biostatsquid

볼케이노 플롯과 유전자 발현 분석에서의 사용법을 소개합니다.

예시 시나리오: 약물을 처리한 두 세포 그룹에서 유전자 발현을 비교하는 경우입니다.

볼케이노 플롯은 약물이 유전자 상향 조절 및 하향 조절에 미치는 영향을 시각화합니다.

볼케이노 플롯은 통계적 유의성(p값)과 변화의 크기(배 변화)를 보여주는 분산형 차트입니다.

Y축: 통계적 유의성(p-값), 가장 유의미한 유전자가 맨 위(가장 낮은 p-값)에 표시됩니다.

X축: 그룹 간 유전자 발현의 차이를 나타내는 배율 변화입니다. 음의 배율 변화는 하향 조절을, 양의 배율 변화는 상향 조절을 나타냅니다.

폴드 변화는 일반적으로 해석과 시각화를 쉽게 하기 위해 로그2로 변환됩니다.

p값과 Y축에서의 표현에 대해 이해합니다.

일반적으로 더 나은 시각화를 위해 음의 로그10 p값이 사용됩니다.

음수 로그10을 사용하면 작은 p값을 관리하기 쉬운 숫자로 변환할 수 있습니다.

유의도가 높은 유전자는 음의 로그10 p값이 더 높습니다.

하향 조절된 유전자는 왼쪽(음의 배율 변화), 상향 조절된 유전자는 오른쪽(양의 배율 변화)에 있습니다.