--- /dev/null
+IMPORTANT
+---------
+
+- consensus output (sto,aln) and compatibilety with bioperl/biopython
+
+- Split Homfam refs in two. Use one part as background HMM, the other parts
+ for benchmarking
+
+- Implement meta flags:
+ accurate: --iterations 3
+ default: --mbed --iterations 1
+ fast: --mbed
+ and for more than 10k sequences: --mbed --mbed-iter
+
+- SSE instructions for hhalign (little use in ClustalO frontend; DD)
+ /
+ Patch new code which already contains SSE instructions
+
+ also fix automake/configure then
+
+- Multi-HMMs; also for Pfam+iteration (FS)
+
+- Show degradation of alignment quality when using x reference sequences
+ added to y random/false sequences
+ (Lio Pachter)
+
+- Seed pre-alignment with M-Coffee, MSAProbs, ...
+
+- GUI/API:
+ Will have to catch/override exits from source.
+ find . -name \*.c -or -name \*.cpp -or -name \*.h | xargs grep 'exit('
+ Also best to allow for user override of
+ void Fatal(char *msg, ...);
+ void Error(char *msg, ...);
+ void Warn(char *msg, ...);
+ void Info(int level, char *msg, ...);
+
+
+- Soeding: DNA/RNA alignment incl. reading of nucleotide HMMs
+
+- Automatic HMM-selection/search/download for input
+
+- Structure input: Psipred predictions are apparently part of their
+ hhms and should therefore be ready to use (part automatic
+ HMM-selection)