Re: Excel ate my DNA!

[Nick Lamb <njl98r ecs soton ac uk>]

On Fri, Jul 16, 2004 at 01:59:28PM +0100, david davemalcolm demon co uk wrote:
> In case you haven't seen this:
> http://www.theregister.co.uk/2004/07/16/excel_vanishing_dna/
>
> Does Gnumeric suffer from similar problems?  OO.org Calc?

Yes, in every case more or less the same problem occurs and the same fix is
necessary. The user must explicitly indicate that the data being entered or
imported is meaningless text, rather than allowing the software to decide
for itself.

The article's authors may be suffering from wishful thinking when they ask
that thousands (millions?) of ordinary users should be inconvenienced by
the need to manually select data types for all input, just so that the
biotech community can continue to enter "cute" names like DEC1 for genes.

It also looks as though the researchers have rather over-estimated the
damage done. If RIKEN clone identifiers are written as described (a quick
look on the web suggests this is right) then a relatively simple bit of
scripting can find the likely original identifier with 99%+ accuracy.

Nick.