Non-parametric and semi-parametric support estimation using SEquential RESampling random walks on biomolecular sequences

Table 3 Medium-gap-length model conditions: estimated alignment statistics

Model condition	Est align length	SP-FN	SP-FP
*MAFFT*
10.A	1552.3	0.294	0.341
10.B	1563.5	0.483	0.533
10.C	1554.0	0.657	0.684
10.D	1507.5	0.747	0.752
10.E	1612.8	0.945	0.943
50.A	1785.7	0.086	0.088
50.B	1714.2	0.105	0.102
50.C	1703.1	0.245	0.230
50.D	1712.2	0.455	0.419
50.E	2319.2	0.963	0.948

Model condition	Est align length	SP-FN	SP-FP
*ClustalW*
10.A	1208.5	0.497	0.556
10.B	1186.2	0.624	0.684
10.C	1144.8	0.711	0.754
10.D	1105.7	0.756	0.786
10.E	1060.1	0.896	0.906

Model condition	Est align length	SP-FN	SP-FP
*FSA*
10.A	2289.3	0.334	0.124
10.B	3418.5	0.585	0.164
10.C	4506.6	0.729	0.211
10.D	5000.9	0.800	0.223
10.E	6657.1	0.907	0.531

The MSA support estimation problem requires an input MSA. MAFFT [9] was used to estimate an input MSA for all model conditions in our study. Our study also included ClustalW [13] and FSA [2] alignments to explore the impact of input alignment quality on downstream support estimation. The following table columns list average statistics for estimated alignments on each model condition (\(n=20\)). “Est align length” is the estimated alignment length. “SP-FN” and “SP-FP” are the proportion of homologies that appear in the true alignment but not in the estimated alignment and vice versa, respectively

ISSN: 1748-7188