The BioMicro Center offers high-throuhgput RNA library preparation (HTL-RNA) for both standard input and low input samples. Together, these high-throughput methodologies provide a comprehensive set of options to address a diverse range of experiments investigating RNA biology. High-throughput library generation focuses on reducing price and processing time on a per sample basis. To do this, there are several key differences from our standard library preparation service. First, HTL-RNA services have set batch sizes (24 and 96) with fixed price points (i.e. charge for 16 samples will be the as 24 samples; similarly for 80 samples as 96). If under these batch sizes, DO MORE REPLICATES to simultaneously enhance experimental power and reduce price per sample. Some samples may fail and the cost of re-prepping those samples is NOT included in the quoted cost. Re-prep is only deemed necessary by the core where we have found evidence of a reagent failure. User request re-preps of individuals samples can be done by hand, but at a higher rate. HTL focuses on reducing price and processing time per sample, so some routine services provided with standard library preparation - such as initial and final quality control for all samples, sample arraying, or automatic re-prep of failed samples - are not built in. These services are available as add ons to the original service request.
Reminder: high throughput projects are treated with one condition (PCR cycles, fragmentation time, etc...). They are meant for allowing high replicates of the same samples. High throughput submissions for projects with multiple sub-projects from separate individuals combined on the same plate result in a low success rate due to the different conditions for each sub project. Therefore, htl submissions from multiple projects will tolerate and may expect, failure rates well over 10%.
STANDARD INPUT RNA-SEQ (Eurkaryotic)
NEB Ultra II Directional RNA with Poly(A) Selection
NEBNext Ultra II Directional RNA preparation with poly(A) selection is used to selectively capture mRNA from high-quality total RNA. The selection of mRNA transcripts occurs by hybridization of the poly(A)-tail to magnetic oligo-d(T) beads, which we have miniaturized to 1/6th the original reaction volume using our Tecan Freedom Evo. The purified RNA is then further processed at a 1/10th reaction volume on the Mosquito HV to reliably generate RNA-Seq libraries with inputs ranging from 10ng to 100ng (recommended input of > 50ng). Because this method is reliant on poly(A)-tails during RNA isolation the sample quality significantly impacts performance, as such high-quality samples are an absolute requirement.
NEBNext Ultra II Directional RNA preparation with ribosomal RNA (rRNA) depletion is used to deplete rRNA by enzymatic degradation using single-stranded DNA probes that target rRNAs, leaving all other RNA species present and available for library preparation. This depletion method is agnostic to input sample quality and is the go-to method if samples don't meet quality requirements for poly(A) selection. We have miniaturized this rRNA depletion method to 1/6th and the following RNA library preparation to 1/10th the original reaction volume on the Mosquito HV. This method is suited for inputs ranging from 5ng to 100ng (recommended input of > 50ng). Due to probe design used in the depletion, this method is currently only available for species where NEB has probes available.
Our Cv2 prep is a "home-brew" version of Clontech's SMART-Seq kits which generates cDNA of full length transcripts from low inputs of RNA (i.e. single cell, dilute RNA sample). This workflow plugs into either our HTL Nextera Flex or XT preps where the cDNA is tagmented and indexed.
This method uses oligo-d(T) priming to selectively process polyadenylated RNA which is then amplified by an initial round of PCR, making this prep ideal low-input high quality RNA or single cells. Due to the nature method, submissions should be coordinated with BMC to schedule sample drop-off of up to 384 samples to begin processing immediately in order to reduce possible loss/degradation of sample.
SMARTer Stranded Total RNA-Seq Kit v2 - Pico Input Mammalian (ZapR) is used for library preparation of low input degraded RNA samples by generating and amplifying cDNA from total RNA, after which the ribosomal cDNA is depleted through targeted enzymatic degradation using ZapR probes while keeping all other cDNA for library preparation. This method is typically used for low concentration RNA samples of poor quality (RIN<7 or DV200<60%). We have miniaturized the ZapR method to 1/10th the original reaction volume on the Mosquito HV. This method is capable of handling inputs ranging from 250pg to 10ng (or 50ng for FFPE) however, performance for a given sample input is dependent on the quality of the sample itself. Due to the design of the ZapR probes, this method is only available for Human samples.
High-Throughput 3' Digital Gene Expression (HT3DGE) uses a combination of molecular tagged indexes, SMARTseq chemistry and Nextera Tagmentation to produce libraries derived from the 3' ends of transcripts.
The protocol is based on Soumillon et al., as part of a collaboration with the KI High-Throughput Screening Core. In limiting the sequence space used by the samples, fewer reads should be required for a good transcriptome.
HT3DGE uses a very early indexing step to tag each sample with a "well ID" and a molecular ID. Once tagged the samples are immediately pooled to minimize costs but failed samples cannot be easily reprepped and are not identifiable until sequencing. As such, the method is best suited for experiments where the NUMBER of samples is not limiting (material can be). The 3' nature of the protocol also limits the utility of this method in splicing applications but it should be more robust to using imperfect RNA (RIN 7+ instead of 9+ for standard RNAseq)
Analysis of HT3DGE is not a standard method supported by most open source platforms. We have built pipelines to work with this data, but working closely with the Bioinformatics team is highly recommended.
Standard 3'DGE workflow.Example 3'DGE Data
USEFUL INFORMATION
RNA Pre-load Submissions
Describes samples that upon submission to the BMC can be immediately plugged into the preparatory method that has been specified in the project submission form. This option is provided to reduce library preparation costs and expedite sample processing by minimizing hands-on time by a BMC staff. The specific requirements are listed for each preparatory method in the tables above. These specifications must be met in order to qualify as a pre-load, submissions otherwise may be subject to additional charges.
Due to the nature of a pre-load submission, the entire volume of sample submitted will be used and as such, additional services are not typically offered unless coordinated with a BMC technician. It is important to keep this in mind when submitting precious samples and the BMC recommends to save a portion of sample when possibly. To avoid unnecessary RNA degradation, all RNA pre-loads should be setup the day of submission and be coordinated with BMC staff. These pre-loads should be prepared with extreme care to avoid possible contamination and reduce freeze thaw cycles to prevent degradation of the samples. Degraded or contaminated samples can have a significant impact on library preparation and sequencing results.
Quadrant Layout
At the BioMicro Center, we organize our high-throughput projects in 384-well plates using a quadrant layout. With quadrant 1 (Q1) representing one 96-well plate, quadrant 2 (Q2) representing a second 96-well plate and so on up to Q4. We ask that plates being submitted start with Q1 and arrayed column-wise. Below are diagrams illustrating quadrant layouts.
Quadrant layout
Quadrant 1 of 384-well plate
Normalization
Normalization is referring to bringing all samples to a uniform concentration. For library preparation, uniform input mass is necessary for proper library generation and as such, calculations for normalization should be based upon concentration in ng/µL. This is in contrast to normalization for pooling of final libraries, where calculations are based upon concentration in nM since the number of molecules is important to achieve a desired read distribution when sequencing.
Unique versus Combinatorial Dual Indexing
Combinatorial dual indexing is a technique that uses a set of Index 1 (denoted as i7) and Index 2(denoted as i5) barcodes that are combined in a manner such that it produces distinct i7 and i5 pairs which increases multiplexing capacity for sequencing. With combinatorial dual indexes, each i7 and i5 is shared among other samples on the same plate, typically with i5's repeating across rows and i7's down columns. These combinations are unique but individual indexes used are not. However, a phenomenon known as 'index hopping' has been observed when sequencing multiplexed libraries that followed single or combinatorial indexing schemes with newer Illumina platforms utilizing ExAmp chemistry such as the NovaSeq (Costello et al., 2018). This swapping of indexes causes reads to be mis-assigned and subsequently excluded from further analysis. The primary strategy employed to mitigate the effects of index hopping is through the utilization of unique dual indexes.
Unique dual indexes (UDI) are non-redundant indexes where each i5 and i7 has a distinct index sequence. As opposed to combinatorial dual indexing, an i7 and i5 index is never repeated nor shared among other samples (i.e. for a 96-well UDI index plate, there are 96 unique i7's and 96 unique i5's). Both the combinations and individual indexes used are unique, and as a result the frequency of mis-assigned reads due to index hopping is greatly reduced. The BioMicro Center recommends that UDI's always be used when sequencing on the NovaSeq. The number of libraries that can be multiplexed and sequenced on a single lane is determined by the total number of UDI's provided for each library preparation method.
