doc: add documentation about post-processing the kernels.dat files

doc/colvars-refman-main.tex

@@ -6812,7 +6812,7 @@
     }
 \end{itemize}
 
-The output file ``\outputName\texttt{.colvars.$<$name$>$.$<$replicaID$>$.kernels.dat}'' includes all deposited uncompressed kernels (the step number of deposition, the kernel center, the Gaussian kernel $\sigma$, the height of the kernel, and the biasing energy in $k_{\mathrm{B}} T$). Besides the \refkey{pmf}{opes_metad|pmf} option, there are two ways to manually estimate the PMF, namely (i) reweighting the trajectories using the biasing energy either in the \texttt{.colvars.traj} file (with \texttt{outputEnergy}) or the \refkey{.misc.traj}{opes_metad|print_trajectory_frequency} file (the \texttt{$<$name$>$.bias} column), and (ii) summing up all the kernels (see Ref.\cite{Invernizzi2020} for more information).
+The output file ``\outputName\texttt{.colvars.$<$name$>$.$<$replicaID$>$.kernels.dat}'' includes all deposited uncompressed kernels (the step number of deposition, the kernel center, the Gaussian kernel $\sigma$, the height of the kernel, and the biasing energy in $k_{\mathrm{B}} T$). The format of \texttt{.kernels.dat} files manages to be compatible with the post-processing tools provided in \href{https://www.plumed.org/doc-v2.7/user-doc/html/opes-metad.html}{the PLUMED OPES tutorial}, so that it is possible to run \texttt{State{\textunderscore}from{\textunderscore}Kernels.py} and then \texttt{FES{\textunderscore}from{\textunderscore}State.py} to get the PMF along the biased CVs. Besides the \refkey{pmf}{opes_metad|pmf} option, there are two ways to manually estimate the PMF, namely (i) reweighting the trajectories using the biasing energy either in the \texttt{.colvars.traj} file (with \texttt{outputEnergy}) or the \refkey{.misc.traj}{opes_metad|print_trajectory_frequency} file (the \texttt{$<$name$>$.bias} column), and (ii) summing up all the kernels (see Ref.\cite{Invernizzi2020} for more information).
 
 \cvsubsubsec{Example input}{sec:colvarbias_opes_example}