There are many programs to align short reads to a reference, and quite a few denovo assemblers as well. The terrific list here gives great links to many tools.
But when I see these nifty toys and pick them up to examine them, it's clear that there's a quietly deadly dangers in many, in fact probably all, of the tools.
Which of these tools are safe to use without fear of patent violations? From my experience in database tools, the patent mindfield is terrible. But it seems even worse in bioinformatics because the field is newer and therefore more of the basic infrastructure has been patented from the start.
The Celera assembler for example, is covered by multiple patents and it's clearly not legal to use their methods, even for academic non-profit use. The GPL license is not what I'm talking about, but the patents underlaying its function. I'm not saying people haven't used the assember, but that their use is likely a violation even if Celera hasn't persued it yet. Celera is also far from the only patent holder.
This applies to almost all of these tools. I doubt any one of them is free from patent restrictions. Even the awesome and elegant Bowtie is likely infringing multiple times, since about anything that uses the BWT to search on DNA sequences is patented.
How are people protecting themselves from this issue? Have patent holders like Celera offered licenses, or indemity, at least for noncommercial users?
How do the researchers of these algorithms defend themselves?
How could any company even think to commercialize such tools unless they have a huge portfolio of patents of their own to play mutual-assured-destruction faceoffs with?
I fear that the answers to these questions is that people have just ignored the issues for now and hope that none of the patent holders decides to start poking at their projects.
But when I see these nifty toys and pick them up to examine them, it's clear that there's a quietly deadly dangers in many, in fact probably all, of the tools.
Which of these tools are safe to use without fear of patent violations? From my experience in database tools, the patent mindfield is terrible. But it seems even worse in bioinformatics because the field is newer and therefore more of the basic infrastructure has been patented from the start.
The Celera assembler for example, is covered by multiple patents and it's clearly not legal to use their methods, even for academic non-profit use. The GPL license is not what I'm talking about, but the patents underlaying its function. I'm not saying people haven't used the assember, but that their use is likely a violation even if Celera hasn't persued it yet. Celera is also far from the only patent holder.
This applies to almost all of these tools. I doubt any one of them is free from patent restrictions. Even the awesome and elegant Bowtie is likely infringing multiple times, since about anything that uses the BWT to search on DNA sequences is patented.
How are people protecting themselves from this issue? Have patent holders like Celera offered licenses, or indemity, at least for noncommercial users?
How do the researchers of these algorithms defend themselves?
How could any company even think to commercialize such tools unless they have a huge portfolio of patents of their own to play mutual-assured-destruction faceoffs with?
I fear that the answers to these questions is that people have just ignored the issues for now and hope that none of the patent holders decides to start poking at their projects.
Comment