I have implemented an mlcs solution using dynamic programming and the results are satisfactory. My problem is that when the number of sequences n increases and those sequences are pretty long, the resulting lcs table is too big and cannot be kept in memory and thus everything crashes.
I have seen solutions here in stack overflow where only the previous rows are kept in memory and while it works for table construction I am unable to do backtracking and get the common subsequence, which is my ultimate goal. I have researched and seen that several optimizations are available for this, specially the Hunt-Szymanski optimization and also solutions based on directed acyclic graphs, DAG.
I am asking now if anyone has a reference to a code, pseudocode that implements those optimized solutions (or others!) as I currently lack the time and the knowledge to do it and I am pretty sure that other people have already solved this. I'll be grateful for any idea, discussion or source!
