Performance Test
Stereo-seq Test Data
We tested a total of 11 samples, including mouse brain and various human tumor samples. Each sample was tested with cellbin, bin20, bin50, and bin100 (if available). The number of genes in bin20 data ranges from about 100 to 900.
Cellbin
Bin20
Bin50
Bin100
Number of genes
300~1,000
100-900
600-4,000
1,900+
Number of cells
300,000-400,000
500,000-700,000
80,000-120,000
20,000-30,000
Running Time and Memory Usage
The NeST/hotspot/hdWGCNA algorithms use n_cpus to control the number of parallel threads. This performance test uses n_cpus=8.
Running time and memory requirements are positively correlated with the number of cells and genes, i.e., the larger the data scale, the higher the computation time and memory usage.
Running time: hotspot > nest/hdWGCNA
Memory: nest/hdWGCNA > hotspot
Last updated