Genego

Gene expression data was analyzed using Array Studio (Omicsoft, V7.2). The data was normalized and a MAS5 report was generated for QC assessment. The ArrayStudio (V7.2), Ingenuity Pathway Analysis (IPA, http://www.ingenuity.com) and GeneGo (Thomson Reuters, MetaCore version 6.19, build 65960) packages were used to identify differentially expressed genes (pFDR < 0.05; fold change >1.3 and <−1.3) compared with mock condition.

Raw expression data from microarray Human HT-12_v4_0_R1 Illumina platform was obtained for day 12 and was loaded on Bioconductor/R software and a probe level analysis was performed on them. The raw intensity data from the microarray was normalized using the Quantile normalization method, log (base 2) transformed, and scaled the median to 0 of all samples. An unpaired Student’s t test for two class comparisons was used to identify the significantly differentially expressed genes between the tested conditions. Two-sided Student’s t test was performed with confidence level = 0.95 with two variances being treated as unequal. Multiple testing corrections were not performed on the nominal p values. Probes for different genes were considered significantly differentially expressed for those with absolute fold change value (|FC|) ≥ 1.5 and nominal p value ≤ 0.05. Functional enrichment analysis was performed on significant genes using the MetaCore module on GeneGO web portal program from Thomson Reuters™ (NY).

In this study, we did literature search of the NCBI PubMed database in order to manually examine pathways connected with drug response. To identify additional targets for pathway-linked regulation, we used a manually curated commercial database GeneGo (MetaCore package, Thomson Reuters, USA), and the MetaCore pathway analysis tool to visualize molecular interactions between the proteins. The manually curated functional molecular links between the top pathways and IC₅₀ values are shown on Supplementary Dataset 12.

Pathway analysis was performed using Ingenuity Pathways Knowledge Base-v8.8 (Ingenuity Systems, content version 17199142, release date 17/09/2013) using Genes with LIMMA adjusted p-value <0.05 and -0.5>S.I.>0.5 as reference set and assuming direct and indirect relationships. Fisher’s exact test p-value<0.05 was deemed as statistically significant.
For network analysis, first, we built Metacore Networks (Analyze Networks algorithm, v6.13 build 43450, GeneGO, Thomson Reuters) using DMPK gene and Myotonic Dystrophy related genes as inputs for the algorithm (ACE, ATXN1, CDC42BPB, CNBP, CELF1, CAPNS1, DMPK, DMWD, DMD, INSR, CDC42BPA, FXYD1, TNFRSF1B, TNF, TSPAN7, TNF, TNFRSF1B) and looked for genes in our dataset that intersected the top-score sub-networks related to DM. Next, we created a core network of direct interactions amongst our LIMMA significant genes; we also expanded the network by adding genes outside our dataset, whose interaction is curated in MetaCore. We identified smaller network communities that were prioritized by number of genes in our dataset. We used this second approach to analyze our data set without any prior knowledge.