FAQs

General:
How many compounds are in the library, and where do they come from?
Are FDA approved and known bioactive compounds included in the library?
What concentrations of compounds can be used in HTS assays?
What volume of compounds are typically used in HTS assays?
What are the compounds dissolved in?
How are the compounds delivered to assay wells?

Chemicals:
Can I just get compounds/extracts?
How do you acquire/plate/manage compounds?
I have some compounds. Can you help me prepare plates from them? Can they be added to the CCG collection?
Do you screen known drug libraries?
How many compounds should I test?
What controls should I use?
What is the concentration for compounds, substances and controls for primary and DRC screens?
How do I use the structural flags?
What are Lipinski rules?
Why should a basic scientist worry whether their compound meets the Lipinski rules?

Database:
How do I find my assay results?
What do the colors on the plate display mean?
What is Z' (read Z-factor)?
What is pIC50?
How do I search data to determine "hits"?

Equipment:
Do I need to do a screen to be able to use the HTS Lab equipment?
How do I select filters for my screen? What are good probes/labels?
What plates do you use? How much is the volume?
How long does it take to test compounds?

Cost:
How much does it cost to run a screen? Are there discounts available?

General:

How many compounds are in the library, and where do they come from?

We currently have about 56,000 compounds from commercial suppliers (see the Chemicals section for details). We also have natural product extracts from Dr. Sherman's research program (over 10,000).

Are FDA approved and known bioactive compounds included in the library?

A number of FDA approved compounds are included in the CCG collection through our purchase of the Spectrum 2000 Collection. This mainly includes off-patent drugs. In addition, we have a plated array of approximately 450 small molecules called the NIH Clinical Collection that has a history of use in human clinical trials. The collection was assembled by the National Institutes of Health (NIH) through the Molecular Libraries Roadmap Initiative as part of its mission to enable the use of compound screens in biomedical research.

Other collections from MS Discovery are available at MicroSource Discovery Systems.

What concentrations of compounds can be used in HTS assays?

We typically suggest screening at about 10 uM. Lower concentrations may be used in cell-based assays to reduce false positives and to bias detection toward more potent compounds. For "difficult" targets it may be useful to screen with higher concentrations (20-100 uM) but the number of false positive results is usually quite a bit higher resulting in more work and expense in subsequent studies. Dose response studies are typically done up to 100 uM (perhaps 300 uM).

What volume of compounds are typically used in HTS assays?

Assays are usually run in 10-20 ul for biochemical screens and 30-75 ul for cell-based screens.

What are the compounds dissolved in?

DMSO

How are the compounds delivered to assay wells?

For a primary screen on 384-well plates (perferred format) the compounds are delivered by HDR (high density replication) "pin" tool on the Biomek FX liquid handling robot from the library plates to the assay plates (containing liquid) at the rate of one plate/minute at ambient temperature at a preset volume (0.1 to 0.2 ul). For dose response determination, a Span 8 Biomek FX robot with "cherry-picking" capacity is used to transfer specific compounds from library plates and sequentially dilute to desired concentrations. For 96-well plates, the multichannel pod on the Biomek FX can transfer 1-2 ul of compounds to plates with or without liquid in the wells.

- back to top -

Chemicals:

Can I just get compounds/extracts?

You can get plates of compounds prepared by the CCG Staff and CCG Robotics equipment (see Recharge Rates for fee schedule). This would permit you to run small screens yourself if you have the necessary equipment. We do not provide individual compounds from our libraries as the amounts we have are only sufficient for screening and nearly all are available commercially—check the suppliers' web sites or contact High Throughput Screening Lab (hts-info@umich.edu) for more information. We do run follow-up dose response studies (1 or 2) on hits from screens. After that it is up to the lab to obtain additional compound for follow-up studies.

If you are interested in ordering a compound from a commercial library, get the Catalog No. and the name of the Vendor for the compound. This number is available under the 'Lot Information' tab whenever you view the details for the compound. Then go to the vendors's website (click here for the links) and use their product order system to order the compound by using the Catalog No. as the product identifier.

How do you acquire/plate/manage compounds?

Chemical compounds have been purchased (or donated) as diversity sets from commercial suppliers. Substances compound plates (example: Natural products from Dr. David Sherman) contain extracts and/or mixtures of compounds. All compound plates are formatted and stored at -20C in either 96-well or 384-well SBS plates dissolved in DMSO at concentrations appropriate for most assay formats such that 5-15 uM is the final compound concentration with DMSO concentrations <1%. All compound (and assay plates after transfers of compounds) are barcoded for identification and tracking in the database.

I have some compounds. Can you help me prepare plates from them? Can they be added to the CCG collection?

We can prepare dilution plates or daughter plates from compounds that you have (see Recharge Rates for fee schedule). We would be delighted to add compounds that you may have to our collection for screening against targets developed by UM faculty and staff clients. Please contact High Throughput Screening Lab (hts-info@umich.edu). This can only be done, however, if the structures are known and made available. Also at the present time, any compounds included in the CCG collection must be freely available to all CCG users. It is simply too complex to keep track of a variety of restrictions on different compound sets. One exception to the known structure restriction is for natural product mixtures. To make it worthwhile, however, a mechanism must be in place to follow-up on isolation and structure elucidation of hits.

Do you screen known drug libraries?

See Spectrum collection info above.

How many compounds should I test?

The more compounds tested, the better chance you have of identifying the most potent and specific inhibitors. We have successfully identified hits from screens of 2-3,000 but we would recommend screening at least 10,000 compounds if you have sufficient material. Screens much smaller than that become inefficient in use of equipment and personnel time and we cannot provide much of a break in cost for very small screens.

What controls should I use?

To reduce complications due to plate-to-plate variations, we strongly encourage controls on each plate in the screen. The compound libraries are arrayed to fill the middle wells (10 on a 96-well plate and 20 on a 384-well plate) leaving the outer row (or outer 2 rows on 384-well) on each side for the controls. This results in 80 or 320 wells per plate for library compounds and 16 or 32 wells per plate for controls.

In a screen, we typically require that you include both a positive and a negative control because we use those for calculations of % inhibition (or stimulation) in the database. The optimal situation is when there is a known active compound (I'll use inhibitor in this example for simplicity). In that case, your negative control would simply be the biological readout of your preparation without any added modulators (e.g. enzyme activity, gene expression, cell growth). The positive control would be the readout in the presence of the inhibitor (hopefully a sufficient amount to produce nearly 100% inhibition). If there is no known inhibitor, we usually suggest using a preparation without your target as the positive control. While this is contrary to how you might in the context of an inhibitor screen leaving out enzyme and reading the signal from substrate alone is often similar to the addition of an inhibitor (i.e. positive control). The advantage of having a chemical "positive control" is that you can use the same biological material in all wells on the plate and add control reagents just to the outer wells.

We usually try to include at least one positive and one negative control on each row of the plate to detect any potential problems with material additions from the Multidrop, which add a row at a time. On 384-well plates this may leave additional wells, which could be used for alternative controls - e.g. a no enzyme control or untransfected, cell in addition to a known inhibitor. Details should be discussed with CCG staff when the screen is being designed.

What is the concentration for compounds, substances and controls for primary and DRC screens?

See above.

How do I use the structural flags?

The CCG cheminformatics group has implemented several collections of structural flags for the Mscreen database. These are collections of structural queries which are used to flag compounds containing problematic structural motifs (sub-structures). Structural flags can be a very useful tool in the prioritization of hits obtained from screens and in the evaluation of compounds for purchase or synthesis.

Structural flags should not be taken as absolutes, i.e. a flagged compound should not be immediately discarded. There are several reasons for this.

1. Different projects have different tolerances for risk. For example, therapeutics used to treat acute illnesses can have a much greater risk profile than therapeutics used to treat chronic illnesses.
2. An effort is made but it is impossible to anticipate every substituent which could render an otherwise reactive motif, less reactive.
3. Since the same compound can often be represented with more than one structure (resonance, tautomers, ionization, etc.) differences in software and the way structures are entered into databases can result in erroneous flagging of compounds.

The following sets of structural flags have been implemented or are planned for Mscreen.

Black Flags – Currently implemented in Mscreen

The Black Flags are Mscreen’s implementation of the MLSMR Excluded Functionality Filters used to exclude undesirable chemical matter from the NIH Molecular Libraries Small Molecule Repository (MLSMR). Most of the flags are taken verbatim from the “MLSMR Excluded Functionality Filters 200605121510.xls” file. Any variation on a flag is noted in the comments column. http://mlsmr.glpg.com

In general, the advice to users is not to pursue compounds flagged by the Black Flags. However, these flags are just alerts! Their purpose is to alert the user to a potential problem. It is impossible to anticipate every substituent which could render an otherwise reactive motif, less reactive and different investigations have different tolerances for risk. Ultimately it is the user’s decision whether to pursue a flagged compound, but they should do so with deliberation. Pursuing a lower potency compound with no obvious liabilities may have a much greater chance of success than following up on a more potent, flagged compound.

Red Flags – Currently implemented in Mscreen but still under development

The purpose of Mscreen Red Flags is to alert the user to compounds containing structural motifs which generally render the compound unusable for lead drug discovery or even fundamental biochemical investigations. Mscreen Red Flags represent chemical motifs that tend to be highly reactive or highly toxic and generate nonspecific effects which may be misinterpreted as desired activity. If any component of a sample contains one of these motifs it may produce aberrant results. Consequently, the Mscreen Red Flags are designed to be applied against all structural components present in the sample (parent molecules, salts, etc.).

In general, the advice to users is not to pursue compounds flagged by the Mscreen Red Flags. However, these flags are just alerts! Their purpose is to alert the user to a potential problem. It is impossible to anticipate every substituent which could render an otherwise reactive motif, less reactive and different investigations have different tolerances for risk. Ultimately it is the user’s decision whether to pursue a flagged compound, but they should do so with deliberation. Pursuing a lower potency compound with no obvious liabilities may have a much greater chance of success than following up on a more potent, flagged compound.

Yellow Flags – Planned for implemented in Mscreen

The purpose of Mscreen Yellow Flags is to alert the user to compounds containing structural motifs which may cause in vivo toxicity. Mscreen Yellow Flags represent chemical motifs that are generally not reactive by themselves but may be metabolized into reactive compounds causing toxicity or non-specific effects in vivo. The more in vivo like the assay, the more likely these motifs will produce aberrant results. The Mscreen Yellow Flags are designed to be applied against all structural components present in the sample (parent molecules, salts, etc.).

In general, the advice to users is they should pursue compounds flagged by the Mscreen Yellow Flags since they may provide valuable structure-activity information. However, priority should be given to compounds with no obvious liabilities even if they are less potent. As progress is made toward more in vivo applications, efforts should be made to replace the offending Mscreen Yellow Flag motifs with safer alternatives.

What are Lipinski rules?

The Lipinski "Rule of 5" is a widely-used set of empirical criteria developed by Dr. Chris Lipinski at Pfizer that is used to help identify compounds that have pharmaceutical properties making them more likely to be successful clinical candidates. They were derived by looking at the properties of known drugs on the market (see [add link to Lipinski literature reference]). In general a compound that fails more than 2 of the 5 rules is unlikely to be suitable for clinical use (though there are a number of very high profile counter-examples). The "Lipinski Rule of Five" is called thus because the cutoffs for each of four parameters are all close to five or a multiple of five. The rules of 5 are:

1. Molecular Weight <500
2. LogP <5 (logP is a measure of hydrophobicity)
3. Hydrogen Bond donor <5
4. Hydrogen Bond Acceptor <10

Why should a basic scientist worry whether their compound meets the Lipinski rules?

Actually, you don't need to slavishly follow the Lipinski formula. As noted above, many excellent clinically used drugs (e.g. atorvastatin (a.k.a. Lipitor) fails Lipinski rules). Indeed some of the rules such as MW<500 and H-Bond donor and acceptor are primarily important for absorption of the drug into the body and are less relevant for biochemical or cell-based studies.However, drugs with extreme molecular properties such as very high molecular weight or high logP values (e. greater than 5) frequently have non-specific effects on biochemical or cell-based measurements. Such compounds may aggregate or cause non-specific toxicity to cells or denaturation of proteins. While we have tried to avoid such compounds in our collections there are significant numbers of "non-Lipinski" compounds. Furthermore, as you pursue structure-activity studies, you may find that such compounds have activity. As long as you are mindful of potential complications, they may turn out to be very useful reagents.

- back to top -

Database:

How do I find my assay results?

Log In to the CCG database. Go to Primary Screen &rt; View Plates for primary screens and Dose Response &rt; View DRC data for dose responses. You can then select your target from the target dropdown and view the results by selecting the date or 'All Dates.'

What do the colors on the plate display mean?

Primary Screen data are normalized to the average values of negative (0%) and positive (100%) controls and the percent value are color-coded by well (blue to red in increments of 20% activity- inhibition or stimulation). So progressively warmer colors represent higher levels of activity/inhibition (based on the assay type).

What is Z' (read: Z-factor)?

The Z -factor calculation is useful during piloting for quality assessment of assay conditions (Zhang et. al. 1999). To quantitatively rank assay conditions, perform control experiments and calculate Z from the data collected:

Z' = 1 - 3*(SDpos+SDneg)/abs(Avpos-Avneg);

SDpos = SD of all positive control values
SDneg = SD of all negative control values
Avpos = Average of all positive control values
Avneg = Average of all negative control values
abs represents absolute value.

1 < Z < 0.9 an excellent assay
0.9 < Z' <0.7 A good assay
0.7 < Z' < 0.5 Hit selections will benefit significantly from any improvement
0.5 = Z' The absolute minimum recommend for high throughput screening

This table may differ slightly from published recommendations. However it is based on the general experience of researchers in our facility. We commonly observe that screening results rarely achieve the high quality levels seen during piloting using controls.

Z'-factor Reference: Ji-Hu Zhang, Thomas D. Y. Chung and Kevin R. Oldenburg (1999). A simple statistical parameter for use in evaluation and validation of high throughput screening assays. J. Biomol. Screen 4:67-73.

What is pIC50?

pIC50 = -log(IC50). IC50 represents the compound/substance concentration required for 50% inhibition. Thus the larger the pIC50 the more potent the compound. A pIC50 of 4 indicates an IC50 of 10-4. A commonly used cut-off for defining potent compounds is a pIC50 &rt; 6 or a compound with submicromolar potency. We calculate pIC50 as follows:

Assuming that you ran a concentration curve study where X is the logarithm of concentration and Y is the response; Then,

Y=Bottom + (Top-Bottom)/(1+10^((pIC50-X)*HillSlope))

Y starts at Bottom and goes to Top with a sigmoid shape. This is identical to the "four parameter logistic equation."

How do I search Data to determine "hits"?

The simplest way is to go to Primary Screen &rt; Search. Select the Target, Method or Investigator on which you would like to search and then select the cutoff (could be 3 standard deviations, 70% etc). The query returns the list of compounds/substances that qualified the query.

- back to top -

Equipment:

Do I need to do a screen to be able to use the HTS Lab equipment?

CCG/HTS lab has a multimode plate reader, Multidrop liquid handlers, plate washers and pipettors that can be used for assay development. With training by CCG staff this equipment is available for UM researchers on a fee basis. Please contact Martha (mjlarsen@umich.edu, 5-9422) to schedule training. Individual use of the Biomek FX Robot is generally not permitted due to the complexity of the system. Any such use requires advance special arrangement with Dr. Neubig.

How do I select filters for my screen? What are good probes/labels?

The Pherastar plate reader has several filter modules and more can be purchased from BMGLabtech. As part of assay design, the CCG staff will suggest probes that will improve your screen. As a general recommendation on fluorescent probes, higher wavelength, red to far red region, work better for most HTS assays than the shorter wavelengths that have increased interference with compounds. For luminescence, firefly luciferase is preferred over Renilla due to the cost of reagents. Also flash luminescence reagents can't be used, as the Pherastar doesn't have an injector delivery system. We have used Steady-Glo very effectively for luciferase detection. Please consult with CCG staff for optimum probe selection, as many references are available.

What plates do you use? How much is the volume?

Any SBS-compliant plates from 6-well to 384 well can be used when developing the assay. For screening, 384-well plates are recommended and the HTS lab has samples from many manufacturers. Plate selection can be critical for cell-based assays and we have low volume plates for conserving reagents and compounds. Most assays are 5-100ul volumes with median of 30ul. If you are planning a protein/ligand screen, you will need roughly 1.5 liter of target protein at the final screening concentration to test the complete compound collection.

How long does it take to test compounds?

Actual assay screening time (compound testing) is dependent on the type of assay: for biochemical assay formats, 5 days; for cell-based screens, 3 weeks. Assay development can take from 2 weeks to several months depending on the stage of development and the ease of conversion to 384-well SBS plate format. Once your screen has been adapted to HTS, it is only a matter of days before data is uploaded and available for viewing. After application for screening, CCG staff will contact your lab about initial meeting(s) to determine time lines and suitability of the screen for HTS development.

- back to top -

Cost:

How much does it cost to run a screen? Are there discounts available?

Recharge rates are available here.

- back to top -