Online Supplementary Material s7

Online Supplementary Information

Computation of inbreeding coefficient (F)

The inbreeding coefficient (F) of the sample was inferred relying on previously published data, based on the frequency of consanguineous marriages in the Sardinian population in the period 1965-1969.13 We considered this period of reference as appropriate since the years of birth of patients showed a normal distribution with mean=1972 and SD=14.5 years (N=87, p- value=.969).

In order to make F (and therefore q) estimation as precise as possible, we have taken into consideration the area of provenance of the patients (up to the parental generation) when they were born to apparently unrelated parents. This has been possible thanks to the availability of detailed data on the frequency of consanguineous marriages (for any given degree of relationship and for any altitudinal zone) reported in previous studies.13,14

Furthermore, the presence of one patient born to 2nd cousins parents (F=1/64= 0.015625) and of two born to 3rd cousins (F=1/256=0.00390625) was considered in the calculations (table

F computation has been carried out in three steps:

1. we estimated α (F for the general population) for each of the four areas in which

Sardinia can be divided on the basis of altitude, relying on the frequency of

consanguineous marriages (1st and 2nd cousins) in every area (table S1);

2. we computed a general F estimate for the whole set of patients as a weighted mean

based on the relative frequency of individuals coming from the four different areas

(when they were born to apparently unrelated parents) and of individuals born to

parents with a certain degree of consanguinity (when they were born to related

parents, table S2); 3. since detailed data on the frequency of consanguineous marriages other than those

among 1st and 2nd cousins were not available and the F computed in table S2 would

have constituted a clear underestimate of the inbreeding coefficient in the Sardinian

population, we added to this estimate the contribution to general F given by other kind

of consanguineous marriages (uncle-niece/aunt-nephew, 1st cousins once removed,

and multiple consanguinity marriages, as reported by Moroni et al. 197213) for the

period 1965-1969 (table S3). Area Altitude 1st cousins 2nd cousins α estimate plains less than 200 m 0.3 0.44 0.00025625 coastal hills 200-600 m 0.74 1.3 0.00066563 internal hills a 200-600 m 0.69 1.39 0.00064844 mountains over 600 m 1.26 2.15 0.00112344

Table S1. Frequency (%) of 1st and 2nd cousins marriages in the four different altitudinal areas which Sardinia can be divided into (original data from Moroni et al. 197213) and relative α estimates, where α is the average inbreeding coefficient in the general population. α calculation was based on the formula ∑ pi x Fi, where pi and Fi are the frequency and characteristic inbreeding coefficient of consanguineous marriages of the ith degree in the same population.14 a Hill towns/cities located more than 10 km away from the sea.

Area of provenance/degree of relative frequency α/F F contribution relationship of the parents in the samplec plainsa 0.00025625 87/178 0.000125246 coastal hillsa 0.000665625 28/178 0.000104705 internal hillsa 0.000648438 45/178 0.000163931 mountainsa 0.001123438 15/178 0.0000946717 2nd cousinsb 0.015625 1/178 0.0000877809 3rd cousinsb 0.00390625 2/178 0.0000438904 Total 178/178 0.000620225 Table S2. General F estimate for the set of patients based on the contribution of every altitudinal area (for patients born to unrelated parents, α coefficient) or, alternatively, of every degree of relationship between parents (for patients born to consanguineous parents, F coefficient). a Individuals born to unrelated parents, for whom the altitudinal area of provenance was kept into account to compute the corresponding F contribution b Individuals born to related parents, for whom the degree of relationship of the parents was kept into account to compute the corresponding F contribution c Detailed and unambiguous information on the geographical provenance of parents was available only for 178 individuals

degree of consanguinity F contribution of the parents 1st and 2nd cousinsa 0.000620225 uncle-niece/aunt-nephewb 0.0000103764 1st cousins once removedb 0.0000981951 other (multiple consanguinity)b 0.0000485185 Total 0.000777315 Table S3. Final computation of the total inbreeding coefficient for the sample based on the frequency of individuals born to related/unrelated parents in the sample. a F estimate for the sample reported in table S2. b original data from Moroni et al. 197213

Mutational spectrum

allele count relative frequency (%) -441_427del 226 66.47 V1146M 29 8.53 2463delC 22 6.47 213-214delAT 7 2.06 A1018V 6 1.76 G1000R 5 1.47 1512-1513insT 5 1.47 H1069Q 3 0.88 R778W 7 2.06 2304-2305insC 5 1.47 2035delC 3 0.88 G1089V 2 0.59 T993M 2 0.59 V890M 2 0.59 G943S 2 0.59 S921N 2 0.59 G869R 2 0.59 R919W 2 0.59 N1270S 2 0.59 L1043P 1 0.29 S921Q 1 0.29 I748F 1 0.29 3852-3875del24 1 0.29 19-20delCA 1 0.29 2122-8T->G 1 0.29 total 340 100a Table S4. Detailed mutational spectrum of the sample of Sardinian WD patients included in the study (i.e. with detailed and unambiguous information on the geographical provenance of parents, N=178). No significant discrepancies with the mutational spectrum of the whole sample (N=192) was noticed (data not shown). a Slight discrepancies of the total sum of relative allelic frequencies from 100 % (due to rounding of single frequency values) have been ignored