Aims. We report four new 22 GHz H2O water masers found in a Green Bank Telescope search toward 37 nearby objects. Our goal was to find new maser galaxies, active galactic nucleus (AGN) disk masers, and objects where hydroxyl and water maser species coexist. Methods. We observed 37 sources within 250 Mpc that were selected by high X-ray luminosity (L-X > 10(40) W) and high absorbing column density (N-H & 10(22) cm(-2)). Sources included dual or triple AGN and interacting systems. We also searched objects detected in hydroxyl (OH). A catalog of 4038 known H2O (non) detections was assembled to avoid unnecessary reobservations. The final selection consisted of 16 new sources, 13 nondetections to follow up with a factor 10 higher sensitivity, 10 OH masers and 1 deep OH absorber, of which 37 were observed. Results. Water megamasers were detected towards the Sy 2 galaxy 2MFGC 13581, towards the 6 GHz OH absorber NGC 4261 and towards the two 1.6 GHz OH maser sources IRAS 17526 + 3253 and IRAS 20550 + 1656. We set upper limits on 33 nondetections. The detection rate was 25% in OH galaxies and 11% overall. The mean sensitivity was 4 mJy over 24.4 kHz (0.31 km s(-1)) or between 0.1 L-circle dot and 1.0 L-circle dot rms for the distances covered by the source sample. Combined with other searches, a total of 95 objects have now been searched for both OH and H2O masers. Conclusions. The maser features in 2MFGC 13581 are typical of a sub-parsec accretion disk, whereas NGC 4261 likely has jet masers in a masing torus. The NGC 4261 galaxy (3C 270; dusty torus, twin jet) and its masers appear similar to NGC 1052, where continuum seed emission by a twin jet supports masers in the torus. Imaging with very long baseline interferometry is required to determine the masing regions in NGC 4261 and 2MFGC 13581. IRAS 17526 + 3253 has narrow 350 L-circle dot systemic masers, and the tentative 5 sigma detection in IRAS 20550 + 1656 (II Zw 96) strongly resembles massive star formation kilomasers in NGC 2146. The latter two detections increase to eight the number of known "dual-species" objects containing both OH and H2O masers. Further, we found the overall dual-species detection rate (8 in 95) to be of the order of the joint probability of both species independently occurring in the same object (1% lower bound). However, this needs to be verified by a more detailed analysis that accounts for the individual selection criteria of the 95 searched objects. Lastly, we see a lack of H2O kilomasers in OH megamaser objects, which was previously noted. This may be due to sensitivity bias rather than for astrophysical reasons.
