Reading single cell papers made me wonder why researchers use 100 cells, 500 cells, or 1000 cells? What's criterion to call such numbers? How do you know you got the whole picture by using such a number in your paper? "The population average is a lie." However, those heat maps (clustering) are averaged out of single cells. You call "single cell analysis" - but you still cluster those single cells - sounds you average out of "single cells."
Given cancer heterogeneity is evolving, how do you know the number of single cells you use at a time point represent or capture the cancer heterogeneity?
What's your thinking?
Strictly speaking, as you commented, the transplantation of a few number of cells more than 100 as a dilution assay is not appropriate. Theoretically, even a single cell can form tumor tissue, but in my experience, this is quite difficult to show reproducible result depending on the character of the single cell even though you have selceted it with cancer stem cell marker given the heterogeneity of tumor cells. Indeed, researchers often investigate into the minimum number of cancer cells which can form the xenograft tissue, and then compare that number as to reflect the stemness in dilution assay.
The term of "single cell" is analysis of each cell condition for example by flow cytometer or microfluidics.
Identifying 100, 500, or 500 maybe, because of personal decision. To make estimate of population based on possible sampling methods, that statistical methods to understand the population is common.
It's better when we use power analysis to determine how many cells should we use.
I agree population average / mean, also median is a lie, not precise and accurate true value. But if we want to have true value, I do not know, what kind of measurement that have that ability.
http://cran.r-project.org/web/packages/pwr/pwr.pdf
Dr. Sandi Sufiandi, can you clarify your statement? Do you mean using PCA to determine the number of single cells? As defined: "Principal component analysis (PCA) is a statistical procedure that uses an orthogonal transformation to convert a set of observations of possibly correlated variables into a set of values of linearly uncorrelated variables called principal components. The number of principal components is less than or equal to the number of original variables. This transformation is defined in such a way that the first principal component has the largest possible variance (that is, accounts for as much of the variability in the data as possible), and each succeeding component in turn has the highest variance possible under the constraint that it is orthogonal to (i.e., uncorrelated with) the preceding components. The principal components are orthogonal because they are the eigenvectors of the covariance matrix, which is symmetric. PCA is sensitive to the relative scaling of the original variables."
I wonder if you got a high PCA, you still cluster on certain conditions (set of genes of your interest) - still "average out" of your single cells, which is not really for single cells per se.
As Go J Yoshida pointed out, a single cell should be a single cell - in physiology, one cell, one clone, is sufficient to get a tumor. However, all publications thus far, PCA their single cells - doesn't make sense in biology.
We view at least 20 cells over at least 3 separate experiments and see if the error for the measurement is low (i.e. p
Professor Suzanne Scarlata,
Thanks so much for sharing your insight. Can you be more specific? What's the specific condition (gene clustering of how many genes?) you "see if the error for the measurement is low (i.e. p
We only work on cultured cell lines which theoretically should be genetically identical. We usually only probe for one specific behavior or characteristic at a time so that we can use standard statistical tests to see if there are significant differences between various groups. We have publications using this simple approach although we have not published raw data sets.
- Badges
- Science topic
- Similar topics
- Methodology
- Laboratory
- Cellular Biology Techniques