I looked back through old files, and I'm reasonably sure that the correct model tree for that paper is shown below.

The measure of accuracy was simply the standard Robinson and Foulds (symmetric distance) measure (Robinson, D. F., and L. R. Foulds. 1981. Comparison of phylogenetic trees. Math. BioSci. 53:131-147), expressed as a percentage difference from the true (model) tree.  In other words, 100% accurate means a symmetric distance between model and estimate of zero; 0% accurate equates to the maximum possible symmetric distance.  I have no idea what Rannala et al. meant or where they got it. It clearly is not correct, especially since the model tree contains a few polytomies. Their measure only deals with true edges, whereas a 100% accurate reconstruction would have to represent the true polytomies as well (to have a symmetric distance of zero).

Let me know if you have any other questions. Good luck with your work.

David Hillis


        model tree = (((1:0.0011635,2:0.0011635):0.0168703,(3:0.0093077,(4:0.0005817,5:0.0000000):0.0052356):0.0232693):0,(((((((((((6:0.0151251,(((((7:0.0040721,((8:0.0040721,(13:0.0319953,27:0.0226876):0.0063991):0.0040721,((10:0.0005817,11:0.0052356):0.0040721,(15:0.0058173,16:0.0069808):0.0069808):0.0029087):0.0029087):0.0011635,9:0.0011635):0.0023269,(12:0.0145433,28:0.0104712):0.0023269):0.0023269,((20:0.0052356,(((((29:0.0046539,31:0.0302501):0.0029087,30:0.0058173):0.0023269,34:0.0058173):0.0029087,32:0.0040721):0.0034904,33:0.0046539):0.0034904):0.0023269,((((21:0.0098895,(24:0.0063991,25:0.0052356):0.0127981):0.0058173,26:0.0139616):0.0040721,22:0.0116346):0.0063991,23:0.0098895):0.0034904):0.0029087):0.0034904,14:0.0139616):0.0034904):0.0040721,18:0.0133798):0.0046539,17:0.0133798):0.0052356,35:0.0145433):0.0034904,(91:0.0168703,136:0.0127981):0.0069808):0.0029087,(((((((((37:0.0098895,((47:0.0058173,(48:0.0023269,49:0.0029087):0.0046539):0.0029087,52:0.0157068):0.0058173):0.0011635,(44:0.0087260,70:0.0081443):0.0046539):0.0023269,((71:0.0052356,((((((73:0.0104712,(80:0.0069808,82:0.0040721):0.0029087):0.0040721,81:0.0052356):0.0011635,(((74:0,75:0.0029087):0.0034904,(76:0.0034904,((77:0.0040721,79:0.0023269):0.0005817,78:0.0023269):0.0017452):0.0075625):0.0005817,((83:0.0005817,(85:0.0,86:0.0075625):0.0):0.0005817,84:0.0081443):0.0011635):0.0005817):0.0005817,(89:0.0034904,90:0.0):0.0):0.0011635,(87:0.0011635,88:0.0000000):0.0005817):0.0034904,((((((((((101:0.0075625,((102:0.0058173,108:0.0069808):0.0011635,(106:0.0087260,(133:0.0040721,135:0.0040721):0.0058173):0.0011635):0.0005817):0.0023269,((109:0.0133798,111:0.0063991):0.0017452,110:0.0069808):0.0017452):0.0005817,((104:0.0087260,105:0.0168703):0.0017452,((112:0.0046539,113:0.0011635):0.0174520,((149:0.0093077,150:0.0052356):0.0011635,153:0.0046539):0.0040721):0.0017452):0.0029087):0.0011635,(103:0.0127981,107:0.0052356):0.0052356):0.0017452,(((120:0.0063991,(134:0.0087260,152:0.0098895):0.0040721):0.0029087,((151:0.0023269,(155:0.0040721,(156:0.0040721,157:0.0017452):0.0011635):0.0011635):0.0005817,158:0.0063991):0.0029087):0.0005817,(137:0.0058173,165:0.0081443):0.0029087):0.0011635):0.0005817,(((((((114:0.0040721,115:0.0034904):0.0052356,159:0.0069808):0.0046539,163:0.0104712):0.0017452,(118:0.0098895,(160:0.0110529,(161:0.0319953,162:0.0075625):0.0069808):0.0075625):0.0058173):0.0023269,(116:0.0011635,117:0.0000000):0.0063991):0.0023269,((129:0.0110529,130:0.0034904):0.0209424,(131:0.0052356,132:0.0157068):0.0046539):0.0058173):0.0017452,(148:0.0023269,164:0.0052356):0.0023269):0.0011635):0.0005817,(((124:0.0110529,154:0.0063991):0.0029087,(125:0.0122164,(126:0.0034904,127:0.0034904):0.0151251):0.0046539):0.0023269,128:0.0040721):0.0005817):0.0011635,((121:0.0058173,122:0.0040721):0.0017452,123:0.0052356):0.0052356):0.0017452,138:0.0093077):0.0023269,(((139:0.0052356,(142:0.0087260,((144:0.0046539,145:0.0087260):0.0052356,147:0.0104712):0.0058173):0.0023269):0.0017452,140:0.0116346):0.0011635,((141:0.0063991,143:0.0052356):0.0029087,146:0.0151251):0.0017452):0.0023269):0.0017452):0.0017452):0.0017452,((((((((92:0.0191972,93:0.0081443):0.0029087,((94:0.0081443,95:0.0098895):0.0034904,(96:0.0029087,(97:0.0017452,(98:0.0023269,(99:0.0098895,100:0.0069808):0.0029087):0.0005817):0.0023269):0.0040721):0.0040721):0.0058173,189:0.0058173):0.0011635,((175:0.0017452,((177:0.0017452,179:0.0052356):0.0017452,(180:0.0168703,182:0.0180337):0.0034904):0.0011635):0.0017452,((183:0.0046539,184:0.0017452):0.0011635,185:0.0034904):0.0046539):0.0017452):0.0005817,((((169:0.0069808,(173:0.0029087,(178:0.0023269,(((181:0.0058173,193:0.0040721):0.0029087,192:0.0046539):0.0011635,(187:0.0046539,188:0.0122164):0.0011635):0.0017452):0.0005817):0.0011635):0.0005817,174:0.0011635):0.0011635,191:0.0087260):0.0011635,(186:0.0040721,190:0.0069808):0.0005817):0.0011635):0.0011635,(170:0.0075625,171:0.0023269):0.0029087):0.0023269,((((((166:0.0052356,194:0.0023269):0.0040721,195:0.0017452):0.0023269,196:0.0040721):0.0040721,(((202:0.0046539,212:0.0151251):0.0023269,221:0.0063991):0.0005817,((217:0,223:0.0023269):0.0005817,218:0.0005817):0.0017452):0.0005817):0.0011635,((203:0.0058173,(225:0.0034904,226:0.0069808):0.0017452):0.0005817,(227:0.0011635,228:0.0034904):0.0052356):0.0005817):0.0017452,(167:0.0104712,((168:0.0116346,(197:0.0063991,198:0.0069808):0.0023269):0.0029087,((172:0.0075625,((((((((176:0.0046539,201:0.0087260):0.0011635,210:0.0029087):0.0017452,208:0.0005817):0.0011635,207:0.0011635):0.0011635,(205:0.0034904,206:0.0052356):0.0023269):0.0011635,(204:0.0023269,(213:0.0040721,222:0.0378127):0.0034904):0.0034904):0.0005817,199:0.0104712):0.0017452,((((209:0.0046539,211:0.0046539):0.0011635,(216:0.0040721,219:0.0081443):0.0029087):0.0011635,(214:0.0046539,220:0.0157068):0.0029087):0.0011635,215:0.0017452):0.0017452):0.0005817):0.0023269,200:0.0069808):0.0011635):0.0005817):0.0017452):0.0023269):0.0017452,224:0.0029087):0.0011635):0.0011635):0.0023269,(65:0.0023269,66:0.0023269):0.0029087):0.0017452,((((54:0.0087260,119:0.0011635):0.0034904,((55:0.0063991,(56:0.0011635,57:0.0011635):0.0063991):0.0034904,((59:0.0034904,60:0.0046539):0.0017452,(61:0.0011635,62:0.0058173):0.0029087):0.0023269):0.0017452):0.0023269,58:0.0005817):0.0011635,(63:0.0058173,64:0.0075625):0.0063991):0.0029087):0.0011635,72:0.0017452):0.0011635,67:0.0104712):0.0017452,(68:0.0017452,69:0.0029087):0.0040721):0.0023269,(42:0.0052356,43:0.0110529):0.0034904):0.0017452):0.0017452,(45:0.0029087,46:0.0017452):0.0122164):0.0040721,19:0.0180337):0.0029087,(36:0.0186155,((50:0.0069808,53:0.0267597):0.0023269,51:0.0040721):0.0081443):0.0023269):0.0046539,39:0.0145433):0.0058173,((38:0.0005817,40:0.0116346):0.0017452,41:0.0023269):0.0052356):0.0401396):0.0;




Prof. Hillis -

I'm a PhD student in computer science working on reconstructing
evolutionary trees, and I wanted to know where to get a copy of the
228-taxon angiosperm tree that you used for your paper in Nature in
1996 ("Inferring complex phylogenies").

Also, I have a question about how you measured the accuracy of the
tree for that paper.  Rannala et al. (1998) cited that paper as
defining accuracy as the number of true positives plus 0.5 * the
number of polytomies (or false negatives) consistent with a refinement
of the true edges.

However, I didn't see anywhere in your paper where you mentioned that
definition of accuracy.  I believe you just called it "percent
accuracy".  Was this how you measured the accuracy for that paper?
Did you define it in some other paper?

Thanks for any help,
Kelly


--
David M. Hillis
Alfred W. Roark Centennial Professor
Section of Integrative Biology
University of Texas
Austin, TX 78712
USA

Office: 512-471-5792
FAX: 512-471-9651
Lab: 512-471-5661