New Approaches to in silico Design of Epitope-Based Vaccines
New Approaches to in silico Design of Epitope-Based Vaccines
New Approaches to in silico Design of Epitope-Based Vaccines
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
106 APPENDIX B. EPITOPE DISCOVERY<br />
Algorithm B.1: Weighted majority vot<strong>in</strong>g. This vot<strong>in</strong>g is applied <strong>to</strong> resolve conflicts <strong>in</strong><br />
results <strong>of</strong> microarray experiments. <strong>Based</strong> on the number <strong>of</strong> absent, marg<strong>in</strong>al and present assignments<br />
a prote<strong>in</strong> is unambiguously assigned <strong>to</strong> exactly one class. The marg<strong>in</strong>al classifications were weighted<br />
only 2/3 <strong>to</strong> achieve a better separation <strong>in</strong> absent and present, reduc<strong>in</strong>g the number <strong>of</strong> marg<strong>in</strong>al<br />
assignments.<br />
a ⇐ number <strong>of</strong> absent calls<br />
m ⇐ number <strong>of</strong> marg<strong>in</strong>al calls<br />
p ⇐ number <strong>of</strong> present calls<br />
sum ⇐ a + 2<br />
3m + p<br />
label ⇐ unassigned<br />
if a/sum > 2<br />
3<br />
label ⇐ absent<br />
else<br />
if p/sum > 2<br />
3<br />
then<br />
label ⇐ present<br />
else<br />
label ⇐ marg<strong>in</strong>al<br />
endif<br />
endif<br />
then