Table 2.

The distribution of number of rules learnt for different targets using HI^all and HI^seq. HI^all can generally describe patterns using fewer rules. This is expected as it uses more background types of biological knowledge. Note the strange bimodal distribution for HI^seq rules. The reason for this is unknown.

Rule number	HIall	HIseq
0	423	485
1	425	371
2	701	228
3	133	137
4	38	81
5	37	49
6	37	38
7	14	31
8	11	119
9	2	4
10	1	0