John Irwin's blog

Happy Canada Day

141 years young today.

ZINC 8 release today

ZINC 8 became the default version of ZINC today. The old version, ZINC 7, is and will remain available at, as are ZINC 6 and ZINC 5 - at the corresponding addresses - for backward compatibility.

We have tried to balance many competing demands, including demands on my time, to deliver a database that is pragmatic for ligand discovery, particularly for target-based virtual screening.

We thank you for your interest in ZINC. We hope you find it useful.

ZINC is big, and you are busy. To save time, here are some numbers: was down midnight to 7am San Francisco time last night

The machines are back up, and seem to be fine. We apologize for any inconvenience.

About ZINC versions

We currently maintain four ZINC versions: ZINC 5, 6, 7, and 8, released in 2005, 2006, 2007 and 2008 respectively. Generally, ZINC grows each year as we add more compounds, and more suppliers. But it also shrinks as compound stocks become depleted.

We recommend the use of ZINC 8 for all purposes. The only reason we can think of for using old ZINC versions is to access molecules that may have vanished for one reason or another in the latest version.

In particular, the search feature, upload feature, and subset features are best in or only are available for ZINC 8.

searches in ZINC8 now search over 10M compounds

Starting today, searches in ZINC 8 will search over 10M compounds. Previously, substructure searches only searched about 2.5 M compounds in ZINC, because we were not updating our search dictionary.

Thus, if you previously did not find what you were looking for, look again, it may be there!

Please note that searching quotas are in effect, to allow us to offer this service for free to everyone.

This is one more step in the slow-but-coming ZINC 8 release. back to normal after being partially offline Fri & Sat

All services are back on line. If you used over the past 24 hours and had a problem, please try again.

DOCK 6 on Blue Gene

A group at IBM in collaboration with Terry Lang have published a "RedPpaer" through the IBM website:

High Throughput Computing Validation for Drug Discovery Using the DOCK Program on a Massively Parallel System

Multiconf-DOCK available as DOCK contributed code

The "Multiconf-DOCK" package of Miteva and Villoutreix is now available from the DOCK website

This work will appear in BMC Bioinformatics shortly:

Sauton N., D. Lagorce, B.O. Villoutreix, M.A. Miteva. MS-DOCK: Accurate multiple conformation generator and rigid docking protocol for multi-step virtual ligand screening. BMCBioinformatics, 2008, 9:184

We thank Miteva and Villoutreix for contributing their code! downtime this coming Fri 25 / Sat 26 April

We are doing maintenance on our cluster this Friday and Saturday. services may be down starting Friday around 5pm through all day Saturday, local time. Sorry for any inconvenience.

Syndicate content