Update README.md

README.md

@@ -8,13 +8,16 @@
 
 ## Description
 
-```limiric``` is a quality control package for detecting damaged cells in single cell RNA sequencing (scRNA-seq) data. Many existing quality control packages are designed to remove droplets that contain more than one cell (doublets) or that are empty, rather than those that are damaged. Directly addressing which droplets are damaged is most often achieved by setting thresholds for metrics such as average mitochondrial gene expression or UMI counts. These metrics, even when calculated dynamically, are highly variable across samples, tissues of origin, cell types, treatment conditions, and species and are thus associated with a high probablity of false positive filtering. But searching for true damaged droplets in a sample-specific manner is tedious and non-intuitive, resulting in user-defined filtering that lacks reproducibility. 
+```limiric``` is a quality control package for identifying damaged cells in single cell RNA sequencing (scRNA-seq) data. Many existing scRNA-seq quality control packages have highly specialized functions to identify droplets that contain more than one cell or no cells, rather than those that are damaged. Directly addressing which droplets contain damaged cells is often achieved by setting thresholds for metrics such as average mitochondrial gene expression and UMI count. These thresholds, even when dynamically calculated, can vary significantly across samples, tissues of origin, cell types, treatment conditions, and species, resulting in a high probablity of false positive filtering. But searching for true damaged droplets in a sample-specific manner is tedious and non-intuitive and leads to user-defined filtering that lacks reproducibility. 
 
 
-```limiric``` automates the sample-specific detection of damaged cells in one fast-performing and highly-reproducible function. It operates on the basic principle that damaged and healthy cells can be differentiated in lower dimensional space by their mitochondrial and ribosomal gene expression profiles. In addition to predicting damaged cells, ```limiric``` can perform other pre-processing tasks including removing red blood cells, correcting for ambient RNA with the ```SoupX``` package, and isolating immune cells. There is also the option for it to be used in combination with the ```DropletQC``` package which, while not a requirement as measures of spliced and unspliced RNA are not always available, refines ```limiric```'s damaged cell detection. 
+```limiric``` automates the sample-specific detection of damaged cells in one fast acting and highly reproducible function. It operates on the basic principle that damaged and healthy cells are differentiable in lower dimensional space by mitochondrial and ribosomal gene expression and complexity. Here, dimensionality reduction is performed by applying the Louvain clustering algorithm to a shared nearest neighbor graph constructed for each droplet using the full set of genes encoding mitochondrial and ribosomal structure and function.
 
 
-The package was developed around the community standard ```Seurat``` suite and is designed to incorporate seemlessly into a user's pre-existing ```Seurat``` workflow. However, the main output of ```limiric``` is ```Seurat``` independent and can be used in any single cell analysis platform. 
+In addition to predicting damaged cells, ```limiric``` can perform other pre-processing tasks including removing red blood cells, correcting for ambient RNA (```SoupX```), and isolating immune cells. There is also the option for result verification with ```DropletQC```, a community-accepted package currently available for damaged cell detection. This requires additional input of spliced and unspliced RNA counts for each cell barcode. However, generating these measures is tedious to perform alongside alignment of raw data and is often unavailable in publicly available data. While ```DropletQC``` can refine ```limiric```'s damaged cell annotation, ```limiric```'s performance without ```DropletQC``` validation is near identical to that without it.
+
+
+The package was developed around the community standard ```Seurat``` suite and is designed to incorporate seemlessly into a user's pre-existing ```Seurat``` workflow. However, the main output of ```limiric``` is independent of ```Seurat``` and can be used in any single cell analysis platform. 
 
 <br>