Filtered blog posts by category.
05 October / / preprocess / batch effects / t-SNE / markers
Sequencing datasets often suffer from undesired technical variability, causing some cells to pick up more signal than others. The detection rate can vary considerably among genes. Ultimately, the measurements can also vary significantly when comparing data from different runs of the same experiment, taken on a different day, by a different technician, and so forth. This calls for preprocessing and normalization methods, making the values comparable across cells, genes and experimental conditions.