Ercc spike in mix

Single cells were captured on medium sized Fluidigm C1 Single Cell Integrated Fluidic Circuit (IFC). The SMARTer Ultra Low RNA Kit for Illumina (Clontech) was used for single-cell capture (verified via microscope), on-chip lysis, reverse transcription, and complementary DNA (cDNA) generation. An ERCC Spike-in Mix (Ambion) was used as the technical control per Fluidigm recommendation. Single-cell cDNA quantification was performed using Agilent High Sensitivity DNA Kits and Quant-it Picogreen dsDNA (Invitrogen). cDNA was normalized to 0.20ng/μL with Biomek NX^P(Beckman Coulter). The Nextera XT DNA Library Prep Kit (Illumina) was used for dual indexing and amplification following the Fluidigm C1 protocol. Libraries were purified and size selected twice using $0.9\text{x}$ volume of Agencourt AMPure XP beads (Beckman Coulter). Cleaned libraries were quantified with Quant-it Picogreen dsDNA (Invitrogen) and normalized to 0.3ng/µl with Biomek NX^P (Beckman Coulter). Single-cell RNA sequencing libraries were subsequently pooled for 96-plex sequencing. The resulting cDNA libraries were quantified using High Sensitivity DNA Kit (Agilent). The pooled 96x single-cell libraries were sequenced using HiSeq 2500 (Illumina) with paired end 126bp-8bp-8bp-126bp and 14pM-loading concentration with 5% PhiX spike-in.

293T cells were treated for 48 hours with either control or INTS8 targeting siRNAs as described above. Cells were harvested and washed with PBS. Cells were counted, and 1 × 10⁶ cells were resuspended in 1mL of Trizol and spiked with 1 μL of 1:10 diluted ERCC Spike-in Mix (Invitrogen) and RNA was purified. 1 μg of RNA from each sample was diluted in 10uL of water. The input RNA was subjected to RNA-seq library generation using the TruSeq Stranded Total RNA sequencing kit with RiboZero rRNA depletion (Illumina). The final libraries were amplified to 10 cycles and purified using AMPure XP beads (Beckman Coulter). Pooled libraries (10nM) were sequenced (PE150) on the HiSeq platform at Novogene.
Reads were filtered to require a mean quality score ≥ 20, trimmed to 100 nt, and mapped to the hg38 human genome assembly using STAR. Default parameters were used, except that multimappers were randomly assigned (outMultimapperOrder Random), spurious junctions were filtered (outFilterType BySJout), minimum overhang for non-annotated junctions was set to 8 nt (alignSJoverhangMin 8), and non-canonical alignments were removed (outFilterIntronMotifs RemoveNoncanonicalUnannotated).
The total number of RNA-seq reads aligned in the control or INTS8-dep. samples is described below: