Table 3. Rate of correct classification by UNID-proteomic, ID-proteomic, and REP-PCR fingerprint libraries.
| Source group | UNID-Proteomic |
ID-Proteomic |
REP-PCR |
|||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Sewage | Cow | Dog | Pig | Sewage | Cow | Dog | Pig | Sewage | Cow | Dog | Pig | |
| Sewage | 100 | — | — | — | 90.0 | — | 5.0 | 5.0 | 70.0 | 5.0 | 10.0 | 15.0 |
| Cow | — | 100 | — | — | — | 100 | — | — | 10.0 | 75.0 | 15.0 | — |
| Dog | 10.5 | — | 89.5 | — | 5.3 | — | 89.5 | 5.3 | 10.5 | — | 89.5 | — |
| Pig | 7.1 | — | 7.1 | 85.7 | 14.3 | — | — | 85.7 | 7.1 | — | — | 92.9 |
Data shown in rows indicate for each source group the percentage of its isolates being assigned to its own group (i.e. correct classification) and to each of the three other groups (i.e. incorrect classification) (“−” =0%). Data for UNID- and ID-proteomic fingerprints are on the basis of the complete set of consensus spectra (i.e. no filtering; see Fig. 4). The classification of individual isolates is shown in details in Supplementary Table S2.