################################################################
Dataset Information
################################################################

Dataset Name:
Growth rate-corrected (GR) dose-response metrics across a panel of 71 breast cancer cell lines treated with a library of small molecule and antibody perturbagens. Dataset 2 of 4: Calculated dose-response metrics. 

Dataset Description:
Dose-response data were collected for a panel of 73 breast cancer cell lines grown under standard or modified culture conditions and treated with 139 small molecule and antibody perturbagens. A subset of these data was previously described in Heiser et al. (2012) (PMID: <a href="http://www.ncbi.nlm.nih.gov/pubmed/22003129" target = "_blank">22003129</a>) and Daemen et al. (2013) (PMID: <a href="http://www.ncbi.nlm.nih.gov/pubmed/24176112" target = "_blank">24176112</a>). Here, the complete dataset was analyzed using the growth rate inhibition (GR) metrics described in Hafner et al. (2016) (PMID: <a href="http://www.ncbi.nlm.nih.gov/pubmed/27135972" target = "_blank">27135972</a>) to account for differences in growth rates across the cell lines and treatment conditions.

--Data in Package:
20269.txt

--Metadata in Package:
Small_Molecule_Metadata.txt
Antibody_Metadata.txt
Cell_Line_Metadata.txt

################################################################
Center-specific Information
################################################################

Center-specific Name:
HMS_LINCS

Center-specific Dataset ID:
20269

Center-specific Dataset Link:
http://lincs.hms.harvard.edu/db/datasets/20269/

################################################################
Assay Information
################################################################

Assay Protocol:
1. As reported previously in Kuo et al. (2009) (PMID: <a href="http://www.ncbi.nlm.nih.gov/pubmed/20003408" target = "_blank">20003408</a>) and Heiser et al. (2012) (PMID: <a href="http://www.ncbi.nlm.nih.gov/pubmed/22003129" target = "_blank">22003129</a>), cells were plated at proper density (from 1,000 to 15,000 cells/100 L/well, depending on the cell line) in 96-well plates so that they remained in logarithmic growth at the time of the assay. <br />

2. The cells were allowed to attach overnight before being treated in triplicate with a set of nine doses in 1:5 serial dilution per small molecule or antibody perturbagen. <br />

3. Adenosine triphosphate content was measured as a proxy for relative cell count using the CellTiter-Glo<sup></sup> (CTG) Luminescent Cell Viability Assay (Promega, WI, USA) with slight modifications of the manufacturer's protocol. Briefly, the CTG reagent was diluted with phosphate-buffered saline (PBS; 1:1, volume:volume), and the culture media was removed from the 96-well plate prior to adding 50 L per well of the diluted CTG reagent. Luminescence from the assay was recorded using a BIO-TEK FLx800. Background CTG signal was determined by measuring the luminescent signal in wells without cells. CTG-based cell counts were obtained at day 0 from untreated plates grown in parallel to the point of treatment and at day 3 after perturbagen exposure from treated plates (doi: <a href="http://dx.doi.org/10.7303/syn8094063.1" target = "_blank">10.7303/syn8094063.1</a> and the associated <a href="http://lincs.hms.harvard.edu/wordpress/wp-content/uploads/2017/01/reffilepkg.zip">reference file package</a>).<br />

4. Consistent with the methods reported in Hafner et al. (2016) (PMID: <a href="http://www.ncbi.nlm.nih.gov/pubmed/27135972" target = "_blank">27135972</a>), normalized growth rate inhibition (GR) values for each technical replicate of each biological replicate/cell line/perturbagen/concentration combination were calculated according to the following formula: 2^[log2(x(c)/x0)/log2(xctrl/x0)]-1 where x(c) is the background-subtracted average CTG value measured after a given treatment, x0 is the median background-subtracted CTG value from the day 0 untreated plate, and xctrl is the robust mean of the background-subtracted CTG value of the DMSO-treated control wells on the same treated plate (defined as xctrl=mean(x(abs(log10(x)-log10(mean(x)))<1.5)), where x are all DMSO-treated control values). Background-subtracted CTG values below 1, which occurred only for 0.1% of the data, were set to 1. <br />

5. GR values for each biological replicate/condition combination then were calculated as the robust average of the three technical replicates (defined as x(c)= mean(x(abs(log10(x)-log10(mean(x)))<1)), where x are the technical replicates for a given condition).<br />

6. The nominal division rate of each cell line, CtrlNDiv, was defined as log2(xctrl/x0).<br />

7. The dataset was filtered to remove cell lines for which the untreated controls grew less then 23% over the course of the assay (corresponding to CtrlNDiv > 0.3) and to remove treatments involving reagents (cell lines and perturbagens) for which metadata could not be adequately defined. Following filtering, the dataset consisted of relative cell counts, calculated GR values, and calculated division rates (CtrlNDiv) across 8882 cell line/perturbagen combinations including replicates and 4788 unique cell line/perturbagen combinations excluding replicates (HMS LINCS Dataset #20268).<br />

8. For each biological replicate/cell line/perturbagen combination, mean GR values across all tested concentrations were fitted to a sigmoidal curve to extract GR<sub>50</sub>, GR<sub>max</sub>, GR<sub>inf</sub>, Hill coefficient, GR_AOC, GEC<sub>50</sub>, and r<sup>2</sup> values (HMS LINCS Dataset #20269).<br />

9. To extract GR metrics signatures across the perturbagens tested, we selected only the conditions in which the untreated division time was below 80 hours (corresponding to CtrlNDiv > 0.9) and averaged each GR metric across biological replicates (calculated as the geometric mean for GR<sub>50</sub> and GEC<sub>50</sub> and as the arithmetic mean for all other metrics) for 4650 cell line/perturbagen pairs (a median of 88 perturbagens per cell line). To extract signatures across all perturbagens, we evaluated the median, lower quartile, and upper quartile of each GR metric for each perturbagen across all cell lines (HMS LINCS Dataset #20270). To extract signatures across perturbagen classes, perturbagens were grouped by class based on their target, and the median, lower quartile, and upper quartile of each GR metric was evaluated across all cell line/perturbagen pairs for a given perturbagen class (HMS LINCS Dataset #20271).<br />

Date Updated:
2017-03-14

Date Retrieved from Center:
11/13/2015

################################################################
Metadata Information
################################################################

Metadata information regarding the entities used in the experiments is included in the accompanied metadata. A metadata file per entity category is included in the package. For example, the metadata for all the cell lines that were used in the dataset are included in the Cell_Lines_Metadata.txt file.
Descriptions for each metadata field can be found here: http://www.lincsproject.org/data/data-standards/
[/generic/datapointFile]
[/generic/reagents_studied]
