diff --git a/README.md b/README.md
index 5ef893131283cf81b851988d2d150f878c8c5987..223e5664b888588aeba16c07f07fccb2abfeed41 100644
--- a/README.md
+++ b/README.md
@@ -99,7 +99,7 @@ E\left(\text{norm}IP\left(t\right)\right)= E\left(\frac{\text{norm}IP\left(t\rig
 which gives
 
 ```math
-\text{norm}WCE\left(t\right) = WCE_{x}\left(t\right) \sum_{t=1}^{T_x}\frac{IP_{x}\left(t\right)}{WCE_{x}\left(t\right)} \frac{1}{\sum_{t=1}^{T_x}IP_{x}\left(t\right)}
+\text{norm}WCE\left(t\right) = WCE_{x}\left(t\right) \frac{1}{\sum_{t^{'}=1}^{T_x}IP_{x}\left(t^{'}\right)} \sum_{t^{'}=1}^{T_x}\frac{IP_{x}\left(t^{'}\right)}{WCE_{x}\left(t^{'}\right)} 
 ```
 
 Not that all the computation are scaled by the genome size and not at the read number as in [Hu et al.](https://doi.org/10.1093/nar/gkv670), this is also why we add a scaling factor $\beta$ (default to $10^3$).