Modelling and visualizing fine-scale linkage disequilibrium structure

Edwards, David

doi:10.1186/1471-2105-14-179

BMC Bioinformatics

Table 1 The performance of the two algorithms

From: Modelling and visualizing fine-scale linkage disequilibrium structure

Breed	chr	N	p	\|E _s \|	under	over	\|E _sp \|	under	over	t _s	t _f	t _sp	t _fp	width
duroc	1	4249	2863	4762	0	0	4340	0	0	35084	1153	37	37	15
duroc	2	4249	2113	3676	0	3	3339	2	2	18450	920	26	27	10
duroc	3	4249	1535	2873	1	13	2643	4	15	8695	679	20	20	8
duroc	4	4249	2211	4392	0	0	4012	0	0	24005	1007	35	36	11
duroc	5	4249	1255	2357	7	2	2166	8	2	6030	559	16	16	10
duroc	6	4249	1937	3358	0	0	3069	0	0	15396	778	24	24	12
duroc	7	4249	1911	3978	0	3	3621	0	2	19112	966	36	36	12
duroc	8	4249	1694	3136	3	3	2871	3	3	12193	808	24	24	12
duroc	9	4249	1749	3399	3	14	3087	5	17	13904	841	26	27	14
duroc	10	4249	1078	2178	0	0	1980	0	0	3991	482	15	15	12
duroc	11	4249	1145	2493	0	2	2294	0	3	6206	680	19	19	16
duroc	12	4249	973	1857	0	0	1689	0	0	3138	399	13	13	10
duroc	13	4249	2365	3998	0	7	3634	0	2	23112	1019	29	29	9
duroc	14	4249	2252	4037	0	3	3731	0	2	23291	1016	29	29	14
duroc	15	4249	1605	2977	0	0	2717	0	0	10757	707	22	22	11
duroc	16	4249	1148	2249	0	0	2032	0	0	4936	549	16	16	10
duroc	17	4249	941	1856	1	3	1703	1	3	3034	427	12	12	10
duroc	18	4249	792	1402	0	0	1262	0	0	2032	341	9	10	9
landrace	1	1979	4071	7679	0	5	7020	0	4	87941	1295	63	64	12
landrace	2	1979	2153	4058	0	0	3690	0	0	14868	595	28	29	12
landrace	3	1979	1957	4036	1	5	3664	4	3	13841	605	30	31	17
landrace	4	1979	2210	4498	1	1	4111	3	2	16629	634	33	33	12
landrace	5	1979	1489	3368	0	7	3090	1	4	6511	447	24	25	16
landrace	6	1979	2163	4113	0	0	3753	0	0	14664	577	28	28	12
landrace	7	1979	2116	4634	0	0	4238	0	0	16385	666	36	37	13
landrace	8	1979	1836	3989	2	4	3682	4	4	11989	611	28	29	14
landrace	9	1979	2301	4897	0	4	4415	0	5	20077	758	40	40	13
landrace	10	1979	1077	2285	3	11	2074	9	9	2997	334	16	16	14
landrace	11	1979	999	2150	0	0	1954	0	0	2492	289	15	15	11
landrace	12	1979	1098	2573	0	0	2332	0	0	3420	376	21	21	11
landrace	13	1979	2734	5295	0	7	4868	3	6	30046	870	39	40	12
landrace	14	1979	2709	4819	1	4	4454	5	3	26602	673	33	33	17
landrace	15	1979	1666	3507	8	13	3202	10	12	9888	486	25	25	15
landrace	16	1979	1181	2403	0	0	2179	0	0	3747	339	15	16	10
landrace	17	1979	1025	2227	1	3	2017	3	3	2390	279	15	15	11
landrace	18	1979	934	1890	0	2	1721	3	3	1981	256	12	12	9
yorkshire	1	2123	3904	6700	1	3	6112	1	1	65538	1065	51	52	11
yorkshire	2	2123	2202	4311	1	8	3968	3	9	16548	644	32	33	11
yorkshire	3	2123	1994	4440	0	9	4016	1	10	17020	751	39	39	13
yorkshire	4	2123	2246	4359	3	3	3979	4	3	15429	550	32	32	12
yorkshire	5	2123	1525	3145	0	0	2864	0	0	6471	454	22	22	16
yorkshire	6	2123	2045	3876	2	2	3513	2	2	13313	556	27	27	11
yorkshire	7	2123	2363	5342	0	11	4894	4	10	27929	866	44	46	14
yorkshire	8	2123	1868	3887	1	5	3593	1	3	14254	662	27	28	11
yorkshire	9	2123	2119	4174	0	3	3757	1	5	15108	653	31	31	11
yorkshire	10	2123	1193	2518	0	0	2268	0	0	3757	360	19	19	12
yorkshire	11	2123	1269	2851	0	0	2633	0	0	5501	483	20	20	12
yorkshire	12	2123	1061	2287	0	0	2068	0	0	2885	324	17	17	10
yorkshire	13	2123	2759	4935	0	0	4544	0	0	27100	785	35	35	13
yorkshire	14	2123	2482	4732	0	6	4353	1	4	22286	749	34	33	11
yorkshire	15	2123	1669	3321	0	4	3030	2	5	8385	501	24	24	12
yorkshire	16	2123	1207	2575	1	5	2345	2	5	3734	339	17	17	10
yorkshire	17	2123	1176	2446	2	7	2223	3	8	3621	339	16	17	14
yorkshire	18	2123	825	1618	2	6	1445	6	6	1434	215	10	11	9

The table shows the performance of the algorithms applied to the pig data. N and p denote the numbers of observations in the data and number of SNPs available after filtering. The edge sets found in steps 1 and 2 of the standard algorithm are denoted E_s and E_{s
p}, respectively: for the fast algorithm they are denoted E_f and E_{f
p}. The numbers of undershoots and overshoots, i.e. |E_s∖E_f| and |E_f∖E_s| for step 1, and |E_{s
p}∖E_{f
p}| and |E_{f
p}∖E_{s
p}| for step 2, are also shown. The corresponding running times in seconds are denoted t_s, t_f, t_{s
p} and t_{f
p}, respectively. The backward step is much faster than the forward step in both cases (t_{s
p} and t_{f
p}). Overall, the fast algorithm is 23.4 times faster than the standard algorithm, and its inaccuracy, assessed as $\sum (| E_{s} ∖ E_{f} | + | E_{f} ∖ E_{s} |) / \sum (| E_{s} |)$ , is 0.0012. The last column shows the maximum width of $G = (V, E_{s})$ . The computations were run under Redhat Fedora 10 Linux on a Intel i7 four-core 2.93GHz machine with 48 GB RAM.

Back to article page

ISSN: 1471-2105

Contact us

General enquiries: journalsubmissions@springernature.com