diff --git a/README.md b/README.md
index b5a29b8f..e1ce8342 100644
--- a/README.md
+++ b/README.md
@@ -24,11 +24,15 @@ equation (ODE) models.
 parPE offers the following features:
 
 * MPI-based load-balancing of individual simulations
+  (if you don't need this, and don't require a C++ library, consider using
+  the [pypesto](https://pypesto.readthedocs.io/en/latest/) Python package
+  which is more mature and user-friendly)
 * improved load balancing by intermingling multiple optimization runs
   (multi-start local optimization)
-* simple integration with [SBML](http://sbml.org/) models via
+* integration with [SBML](http://sbml.org/) models via
   [AMICI](https://github.com/AMICI-dev/AMICI) and
   [PEtab](https://github.com/PEtab-dev/PEtab)
+  * hierarchical optimization (https://doi.org/10.1093/bioinformatics/btz581)
 * interfaces to [Ipopt](https://www.coin-or.org/Ipopt/),
   [Ceres](http://ceres-solver.org/),
   [FFSQP](https://www.isr.umd.edu/news/news_story.php?id=4088) and
@@ -37,6 +41,12 @@ parPE offers the following features:
 * Good parallel scaling to up to several thousand cores
   (highly problem dependent)
 
+**Note that this library has been developed for specific research questions
+and certain assumptions may not always hold. Use with caution.
+In particular, certain default settings may need adaptation (in particular,
+parallelization settings and AMICI settings such as the sensitivity method).
+PEtab support is patchy. Always verify your simulation results.**
+
 ## Getting started
 
 Although various modules of parPE can be used independently, the most
@@ -109,7 +119,7 @@ Other sample build scripts are provided as `/build*.sh`.
 
 ## Recently tested compilers
 
-* GCC 13.2.0
+* GCC 14.2.0
 * Clang 18.1.3
 
 ## Containers