2000/2001 ABRF MICROARRAY RESEARCH GROUP STUDY: A CURRENT PROFILE OF MICROARRAY LABORATORIES

Transcription

1 2000/2001 ABRF MICROARRAY RESEARCH GROUP STUDY: A CURRENT PROFILE OF MICROARRAY LABORATORIES by George Grills 1, Chandi Griffin 2, Aldo Massimi 1, Kathryn Lilley 3, Kevin Knudtson 4, and James VanEe 5 1 Albert Einstein College of Medicine, Bronx, NY; 2 University of California, San Francisco, CA; 3 Cambridge University, Cambridge, UK; 4 University of Iowa, Iowa City, IA; and 5 Cornell University, Ithaca, NY.

2 INTRODUCTION This is the second survey of microarray laboratories conducted by the ABRF Microarray Research Group. The first survey covered data collected from Dec to Feb This presentation covers data collected from Nov to Jan The surveys were geared to gather information from academic, pharmaceutical, and commercial laboratories that offer microarray technologies as a shared resource. Individual laboratories that have these technologies could participate. Data from manufacturers of microarray related products was not included in the survey analyses.

3 Number of Labs GeneChip Arrays Spotted Arrays General Survey Year Figure 1. Participation of the ABRF Mircoarray Research Group (MARG) 2000 and 2001 surveys.

4 Number of Labs Pharmaceutical CommercIal Academic Figure 2. Type of institution of those responding to the ABRF Mircoarray Research Group (MARG) 2000 and 2001 surveys.

5 Number of Labs South America Australia/ New Zealand Asia Europe USA/ Canada Figure 3. Geographical location of those responding to the ABRF Mircoarray Research Group (MARG) 2000 and 2001 surveys.

6 Number of Labs GeneChip Spotted 0 < 6 months 6-12 months 1-2 years 2-3 years 3-4 years Years Operational Figure 4. Length of time the respondents laboratories have been using microarrays.

7 18 16 Mean ± SEM = 4.0 ± 0.7 staff 14 Number of Labs one two three four five six+ Number of Staff Min = 0.1, Max = 50, n = 77 83% of labs expect staff to increase over next year Figure 5. Number of personnel in microarray laboratories.

8 Average Number Of Years Of Experience In Microarray Field Staff Mean ± SEM = 1.46 ± 0.12 years Min = 0 yrs Max = 11 yrs n = 75 Facility Director Mean ± SEM = 2.29 ± 0.16 years Min = 0.5 yrs Max = 6 yrs n = 78

9 % of Respondents 90% 80% 70% 60% 50% 40% 30% 20% 10% GeneChip Spotted 0% Specific Technology Convenience & Speed System Trials Recommendation Initial System Cost Array Costs System Flexibility Figure 6. Factors respondents considered when selecting their microarray platform.

10 Spotfire 12% Other 17% Affy_suite 12% Affy_DMT 4% Affy_MicroDB 3% GeneSpring 10% Cluster 14% ScanAlyze 9% TreeView 15% Arrayviewer 4% Figure 7. Microarray analysis software programs used by the respondents.

11 60 50 Number of Labs Throughput Hardware Reliability Software Bioinformatics Funding Research Commitments Figure 8. Areas in which microarray facilities are currently encountering challenges

12 % of Respondents 80% 70% 60% 50% 40% 30% 20% 10% 0% < 6 months 6-12 months 1-2 years GeneChip Spotted Figure 9. Time from installation to acquire satisfactory experimental data.

13 80% 70% GeneChip Spotted % of Respondents 60% 50% 40% 30% 20% 10% 0% Average Intensity Spikes (Exogenous Controls) Other Methods 70% GeneChip 60% Spotted % of Respondents 50% 40% 30% 20% 10% 0% Actin GAPDH Tubulin Spikes Other Figure 10. Approaches use to normalize (top) data and the genes used as reference standards (bottom).

14 100% GeneChip Spotted Average Percentage 80% 60% 40% 20% 0% Labs doing replicate experiments Frequency of replicate experiments per lab Figure 11. Frequency of laboratories using replicates in their microarray experiments.

15 60% 50% GeneChip Spotted % of Respondents 40% 30% 20% 10% 0% 1.5 fold 2.0 fold >2.5 fold Figure 12. Perceived sensitivity of the microarray system. Smallest fold-change users believe their microarray systems can accurately discriminate.

16 Typical Profile of a Laboratory Using the Affymetrix GeneChip Expression System (n = 17) 1 Fluidics Station (68% of labs) 1 GeneChip Scanner (86% of labs) 2 Computer Workstations (54% of labs) Routinely use Test arrays (71% of labs) Throughput = chips/year (mean SD) Satisfactory data in <6 months from system installation (76% of labs)

17 100% 90% 80% % of Respondents 70% 60% 50% 40% 30% 20% 10% 0% RNA Isolation IVT Sample Prep Hybridization Scanning Data Analysis Figure 13. Services offered by laboratories using the Affymetrix GeneChip Expression System.

18 % of Respondents 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 2000 Survey 2001 Survey Figure 14. Percentage of respondents indicating that they usually Test arrays (Test1 or Test2) in their experiments.

19 16 14 Number of Labs Human Mouse Rat Yeast E. Coli Arabidopsis p53 HuSNP Custom Rat 8% Arabidopsis 1% E. Coli >1% Yeast 9% p53 2% HuSNP 2% Human 44% Mouse 34% Figure 15. Types GeneChip arrays used by the responding laboratories (top,n = 17) and the percentage of use of each (bottom, n = 2056).

20 Scanner Array quality Data quality Tech support Fluidics Software Documentation Array cost Highest rating Above average Average Below average Lowest rating 0% 20% 40% 60% 80% 100% % of Respondents % of Respondents 80% 70% 60% 50% 40% 30% 20% 10% 2000 Survey 2001 Survey 0% Very Satisfied Moderately Satisfied Mildly Dissatisfied Very Dissatisfied Figure 16. User evaluation of the Affymetrix GeneChip System by component (top) and overall satisfaction (bottom).

21 Conclusions from the Affymetrix GeneChip Section of the Survey Implementation of GeneChip technology into facility laboratories began about two years ago. The majority of laboratories are moderately satisfied with the overall performance of their GeneChip systems. The GeneChip system configuration currently used in most laboratories consists of one fluidics workstation, one confocal laser scanner and two computers. Most Affymetrix gene expression studies are currently performed with murine and human GeneChip arrays.

22 Profile of a Custom Microarray Facility Utilizes an arrayer, scanner and dedicated computing equipment. Charges per array. Produces 117 arrays per month (range = 1-800). Provides arrays for 6 research groups (range = 1-35). Use fluorescently labelled target DNA (94% of labs). Use cy3 and cy5 as the fluorescent label. Use poly L-lysine and 3- aminopropyltriethoxysilane coated slides. Use protocols other than manufacturers recommendations.

23 Number of Labs Construct Libraries Probe Prep Arraying mrna Isolation Target Prep Hybridization Scanning Figure 17. Types of services offered by a custom microarray facility (n = 49 labs).

24 $350 Service Fee (avg + SD) $300 $250 $200 $150 $100 $50 $0 Standard array Custom array Hybridize and scan Scanning Only Plasmid prep Sample prep [labeling] Figure 18. Fees for services offered by a custom microarray facility.

25 Robbins Scientific 22% Tecan 9% Beckman 35% Qiagen 17% MWG 4% Hamilton 4% CCS Packard 9% Figure 19. Liquid handling systems used by custom microarray facilities (n = 29).

26 Unspecified Norgren 5% Industries 4% In-House/ Pat Brown 17% Promedia Associates 2% Genetix 2% Engineering Services 2% Cartesian Affymetrix/GMS 4% 23% Biorobotics 4% Intelligent Bio-Systems 2% GeneMachines 16% Genomic Solutions 2% ESI 4% BioRobotics 9% Hitachi 4% Figure 20. Brands of slide arraying instruments used by custom microarray facilities (n = 55). The most widely used printers are shown in blue.

27 Virtek Vision 2% Promedia Associates 2% Molecular Dynamics 2% Unspecified 11% In-House 2% Affymetrix/GMS 13% Genomic Solutions 3% Agilent/HP 5% Biorad 2% Axon 24% Packard/GSI Lumonics 34% Figure 21. Brands of scanners used by custom microarray facilities (n = 63). The more widely purchased scanners are shown in blue.

28 bacterial 7% yeast 6% plant 6% oligos 22% viral 3% rat 9% fly 3% protein 3% other 6% antibody 2% clones 1% human 30% mouse 29% PCR product 45% cdna 22% Plasmids 5% Figure 22. Source (top) and substrate type (bottom) of material spotted by microarray facilities.

29 Table 1. Microarray facility costs. Costs to prepare cdna and oligonucleotide substrates, acquire instrumentation and reagents 1, and label targets 2. cdna Characteristic Average ± SD Maximum Minimum Length of cdnas spotted on array 1039 ± 421 bases 4000 bases 50 bases Density of features per slide 7154 ± Size of features 150 ± 50 m 250 m 50 m Distance between features 219 ± 104 m 550 m 75 m Cost per feature $0.86 ± $1.33 $2.66 $0.003 Oligos Characteristic Average ± SD Maximum Minimum Length of oligos spotted on array 37 ± 20 bases 70 bases 14 bases Density of features per slide 4350 ± , Size of features 134 ± 49 m 250 m 90 m Distance between features 256 ± 77 m 500 m 150 m Cost per feature $2.64 ± $5.25 N/A N/A 1 Facility set up = $286,000 $162,000 (range = $20K - $700K) 2 Preparation of labeled target material = $82 $20 per target (range = $ $300)

30 % of Respondents 60% 50% 40% 30% 20% 10% Very satisfied Moderately satisfied Mildly dissatisfied Very dissatisfied 0% Scanner performance Scanner support Arrayer performance Arrayer support 90% 80% % of Respondents 70% 60% 50% 40% 30% 20% 10% Found reliable Would purchase again 0% Arrayer Scanner Figure 23. User satisfaction (top) with and reliability (bottom) of their custom microarray instrumentation. Note: These data were not broken down by manufacturer.

31 Summary of the Custom Microarray Section of the Survey The cdna arrayers use a wide range of instrumentation from different manufacturers. All have a similar slide format derived from Pat Brown s concept. The predominance of some cdna microarray related companies may be correlated to the length of time they have been in the market.

32 New Users Current Users Expanding Use Number of Labs Affy GeneChip Non-Affy Oligo Arrays cdna Arrays Peptide Arrays Tissue Arrays Membrane Arrays Figure 24. Future directions of microarray laboratories. Most (83%) microarray laboratories plan to expand both staff and instrumentation, and both custom arrayer and Affymetrix users indicate a trend to create shared resource laboratories in which these technologies coexist and complement each other.