Version information

This is version 1.0@457889d of the GPP Web Portal.

Frequently Asked Questions

For frequently asked questions about our different software packages and tools, please refer to their respective pages on our site:

Material Requests

Q: Can I get shRNA clones from you?

Thank you for your interest in the TRC library. However, the Genetic Perturbation Platform at the Broad Institute does not provide TRC clones directly outside of the Broad community. Sigma distributes the glycerol, DNA, and virus of both pLKO.1 and pLKO_005 libraries, whereas Thermo Scientific (formerly Open Biosystems) only distributes the glycerol of pLKO.1 library. The TRC Library is available through Sigma: and Thermo Scientific:

Q: Can I get ORF clones from you?

The ways of getting our ORF clones were addressed in the Supplementary Note 1 of Yang et al. Nat Methods. 2011 Jun 26;8(8):659-61, as follows:

Supplementary Note 1: Availability of clones and distribution procedures.

All clones described in this manuscript are publically available. All lentiviral Expression Clones and hORFeome V8.1 Entry Clones are available via members of the ORFeome Collaboration (, including the Dana-Farber/Harvard Cancer Center (DF/HCC) DNA Resource Core DNA Repository ( and the DNASU Plasmid Repository at ASU Biodesign Institute ( To receive exact clones described in this manuscript, clone requests must include the unique clone identifier numbers provided in Columns 1-2 of Supplementary Table 4 (e.g. ccsbBroadEn_12345 as an example for a specific entry clone and ccsbBroad304_12345 as an example for a specific expression clone). As indicated in online Methods and Supplementary Note 3, an additional distributor, Addgene (, distributes kinase ORF entry clones, pLX expression vectors and control ORF Entry vectors. Additional public and commercial distribution centers may be added. Please contact David E. Hill ( with any questions regarding clone availability and clone distribution procedures.

Q: Can you send me the pLKO vector?

The pLKO.1 cloning vector is available via Addgene. Plasmid: pLKO.1 - TRC cloning vector ( We would also recommend that you obtain a control clone such as: Plasmid: pLKO.1 - TRC control (

Q: Where can I find your mouse ORF clones?

Currently, our collection only contains human ORF constructs, with no immediate plans to generate a mouse library. However, in many cases, using a homologous human ORF as a substitute in a mouse model will work on a functional level.

Q: What fraction of the human ORF library in vector pLX_317 are perfect matches to human genes?

Summary of the human ORF library:

Note that some genes have multiple clones and/or isoforms thus resulting in some of the clone-to-gene discrepancy listed above.

Q: How do I cite usage of the human ORF library?

Yang et al, Nat. Methods 2011


Q: Do you have protocols for lab methods?

Our protocols are available on our protocols page.

Q: Do you have a protocol for Illumina deconvolution of pooled screens?

Yes, please find these on our protocols page, under the heading "Pooled Screening Protocols".

Q: Do you have an infection protocol for a particular cell line?

While we have experience with many different cell lines and assays, we do not provide cell line specific infection protocols as they would vary greatly depending on the intended downstream assay. Generally, our advice is to find your optimal assay conditions (time point, cell density, etc.) and work backwards to find the infection conditions that lead to that.

Design and Scoring

Q: What is the difference between an shRNA targeting the CDS or the 3'UTR region?

We have found that hairpins targeting the first part of the 3'UTR can be just as effective as CDS-targeting hairpins. Thus there is no a priori reason to avoid using the 3'UTR hairpins listed for this gene.

Q: What is the difference between the intrinsic and adjusted scores for an shRNA?

The intrinsic score (also called "original score") of an shRNA assesses the target sequence for both predicted cloneability and knockdown performance, using an evolving set of rules. The scores reported in this website are calculated according to our most current rule set ("Rule Set 9"), and have a maximum score of 15 for a hairpin with the best possible predicted cloneability and knockdown performance and a minimum of 0, for hairpins which exhibit sequence characteristics which could prohibit efficient cloning and/or degrade knockdown performance. We do not produce new clones with 0 intrinsic scores, but as our design rules have changed with time, there are older clones in the library which now have this score. The intrinsic score is not taxon- or target gene-specific, but is based solely on the sequence of the hairpin.

The adjusted score (also called "specificity score") is a function of the intrinsic score and two other factors which take into account the transcriptome of the target organism. The most important of these is the specificity factor, which penalizes hairpins with off-target matches (i.e. to genes other than the originally targeted gene). In absence of any off-target matches (specificity factor values greater than 1), the score is also increased slightly for hairpins which match the majority of the transcript isoforms of the target gene. The other contributor to the adjusted score is the miRSeed factor, which assesses the likelihood that the hairpin's microRNA seed region will exhibit unwanted microRNA effects, either due to sequence similarity with known miRNAs from the target organism, or to particularly common 7mer sequences in the 3' UTR regions of the organism's transcriptome. These factors are multiplied together with the intrinsic score to produce the final adjusted score, and again, higher scores are better. The specificity factor itself is generally not shown separately, except in the candidate hairpin design page for a given transcript, in which hairpin designs with specificity factor > 1 are given higher rank regardless of adjusted score.

Q: Why are there varying numbers of T's in the T runs at the end of the antisense strand of TRC hairpins? Some have 4, some have 5 and some have 6 T's.

Over the course of TRC history, there have been a few changes in hairpin designs and the methods to produce clones. In terms of T runs, we started out designing 5 Ts as a Pol3 termination signal (4 Ts should do, but we added an extra). We produced two-thirds of the TRC1 library using oligos synthesized by IDT in a plate format. IDT oligos have good fidelity, and with this method we only accepted clones recovered with strictly 5 Ts.

Later we implemented a pooled cloning strategy, which uses an oligo pool (thousands of oligos in a pool) that was synthesized on a chip. This synthesis process has lower fidelity than IDT oligos, particularly in a run of mono-nucleotides. Our tests observed deletions in the run of Ts. So instead of 5 Ts, we put 6 Ts in each pooled oligo sequence. When hairpins were fully sequenced, we would still only accept clones which had perfect hairpin stem sequences, but as for the run of Ts, we accepted clones with either 5 or 6 Ts as good clones. So with this method, the run of Ts varies, with the majority at 6 Ts. While we do not yet explicitly annotate this difference among clones in the database, we will soon be doing so.

Knockdown Data

Q: Do you have Knockdown (KD) validation data available for the library?

We do not have KD data that is publicly available yet. We are working on collecting the data and, once published, it will become available to the public through the website.

Q: Is knockdown of a transcript still possible if the score is lower than 1?

Knockdown is definitely still possible with scores below 1--our prediction rules are far from perfect. Clones in the pLKO_005 vector tend to have higher scores because these clones were designed more recently, using newer versions of our design rules. Older clones may have scored well using the prediction rules at the time, but may not look so good under the latest (reported) scoring system.

Vectors and Controls

Q: Has the Genetic Perturbation Platform tested the replication competency of its lentiviral particles post-packaging?

The replication incompetency is a characteristic of the viral vector and packaging system used. Specifically, we use a 2nd generation transfer vector system with 3 vectors: psPAX2, VSVG and a self-inactivating (SIN) viral vector containing a modified 3'LTR. For more details regarding this system, please see the description on Addgene's website. Vector sequences and maps can be found on this website. These systems have been tested at the Broad Institute for production of replication competent virus (RCV) without finding any.

Q: What is the difference between the pLKO.1 and TRC005 (e.g. pLKO.5) vectors?

There are 2 main differences between the vectors, but the overall titer and function are similar.

  1. The TRC005 vector has been engineered with modular components flanked by unique restriction sites to aid molecular alteration of the backbone for other purposes (eg. alternate selection markers, or fluorescent protein expression, alternate promoters, etc).
  2. TRC005 contains a WPRE element 3' to the PGK-PAC cassette. This may have a slight improvement on titer and/or expression from the pol II promoter.

Q: Can I transfer hairpins from pLKO into pol II driven vectors and use your validation or scoring data?

We do not know if knockdown performance of an shRNA sequence in the pLKO vectors is correlated with performance in other vectors, either vectors that drive the shRNA expression with a polIII promoter (pLKO vectors), or that express longer hairpin-containing sequences from a polII promoter (pTRIPZ system). Likewise, our scoring system for shRNAs is based only on knockdown data in the pLKO vectors. We do not know if these scores are correlated to performance in other vector contexts. The few anecdotes we know of using the same shRNA sequence are insufficient to draw any general conclusions one way or the other.

Q: What are the differences between DONR/Entry/Destination/Expression vectors and clones?

DONR and Destination vectors have the ccdb/Cmr cassette in them, also known as the Gateway cassette, and thus must be grown in ccdb-resistant bacteria at 30 ° C. Entry clones require kan or spec, based on what pDONR was used; almost all of our Entry clones require spec. For the expression vectors we use, they are almost always grown in carbenicillin.

Q: Is there a simple way of including a tag in gateway vectors?

While many of our vectors already have a tag, the ultimate answer will depend on the exact vector. If the vector contains useful restriction sites adjacent to the gateway cassette, then yes. If not, more advanced molecular cloning methods are required.

Q: How do I get scrambled negative controls?

We use controls targeting reporter genes or empty vector rather than scrambled. These are available through Thermo Scientific and Sigma.


Q: What are the differences between the chRNAs used in your vectors?

Our vector collection utilizes a few different chRNAs, as follows:

Q: What are the differences between Cas9 and Cas9v2?

Cas9v2 differs in nucleotide sequence (codon usage) from some of our initial Cas9 vectors. Cas9v2 also differs in the identity and position of epitope tags and nuclear localization signals (NLS). There is currently no evidence that Cas9v2 performs significantly different from Cas9.

Q: What are the differences between the EFS and EF1a promoters?

The EFS promoter contains the first ~250 nucleotides of the full EF1a promoter. The full EF1a promoter includes an intron that is spliced out, which is thought to result in higher expression of the downstream elements.

Q: How do I cite usage of the Brunello (human whole genome) sgRNA library?

Doench et al., Nat. Biotechnol., 2016