Re: What skills are necessary for starting a bioinformatics career?

Migrating to bioinformatics? You are probably not alone in thinking that 
this would be an interesting career move. And from your self-description I 
would suggest than you are already well along the required path. 

Ideally, what is needed are people with experience in both computational 
areas (programming, computer science, database design) and in some aspect 
of the biological sciences (experimental biology, biochemistry, genetics, 
structural biology, and the like). Such individuals are not terribly 
common, but neither are they totally non-existent. Most of the 
now "standard" bioinformatics software (BLAST, HMMER, GCG, Staden, etc.) 
and molecular modeling tools (Rasmol, WHATIF, Threader, etc.) were in fact 
created by biological scientists who enjoyed(?) programming. I would guess 
that even the few programs mentioned above are written in more than only 1-
2 computer languages. Much is still done in C or C++, but I think the 
majority of "programming" done in bioinformatics presently is done in Perl.
It has extremely powerful text-handling abilities, and much much more, so 
I would feel remiss to not encourage you (and others) to learn it. Object-
oriented programming is currently quite popular, and likely to remain that 
way for at least the near future. Some familiarity with SQL and modern 
database engines (SYBASE, Oracle, whatever) is increasingly sought-after.
By the way, I hear that Java is also quite useful (and fun).

As for databases, there are probably well over 100 now, and counting. No 
one is intimately familar with them all, but each year the first issue of 
the journal Nucleic Acids Research is devoted to status reports of old & 
new biological databases. Some familiarity with the major sequence 
databases -- GenBank, dbEST, Swissprot, EMBL, PIR, DDBJ,... is already a 
good place to start. For structures, the latest incarnation of PDB (now 
part of the RCSB) is a must, and there are numerous places to go for 
classifications by domain (PFAM, SMART, PRODOM,...) or by structure (SCOP, 
CATH, ...) etc. 

It would also be advantageous to have some familiarity with the present 
state-of-the-art analysis tools, such as WU-BLAST2, PSI-BLAST, SAM, HMMER, 
GCG, CLUSTALW, THREADER. I could easily go on and on...

In the end, there is no right or wrong list of things required. It all 
depends on your previous background, current & future interests, and what 
is easily available to you. Formal training is becoming easier to find, 
but hands-on experience is also important. And there are now several good 
texts devoted to bioinformatics -- at various levels of depth. Forgive me 
for mentioning one in particular: Computational Methods in Molecular 
Biology (see http://www.cs.jhu.edu/~salzberg/compbio-book.html), for 
rather personal reasons. But many others exist which are equally useful.

I wish you a successful, and also an interesting and enjoyable journey 
toward bioinformatics.