Recording guides so an experiment can be repeated

The freezer box says "guide 3" in a handwriting that belongs to someone who has left. Recording guides so an experiment can be repeated is the page that person should have filled before they left. A repeat means another scientist can order the same RNA or regrow the same plasmid and aim it at the same bases, without a private memory of which strand was which. The biology those bases are meant to trigger is in how CRISPR-Cas9 editing works in research. The evidence that the aim worked is in checking whether a genome edit worked.

The spacer, in one direction

Write the spacer 5-prime to 3-prime as the targeting sequence in the guide RNA, not as the DNA strand you happened to copy from a browser. Label the direction on the line. Include the length. SpCas9 spacers and Cas12a spacers are not the same length by default, and a truncated paste is a different guide.

If the experiment used a separate tracrRNA, record that RNA too, or the product class, not only the spacer. A single-guide fusion and a two-RNA anneal are different tubes. The glossary entry is in a glossary of genome-editing terms.

Do not bury the spacer inside a paragraph of cloning steps. Put it on its own line where a search can find it. A screenshot of a design tool is useful only if the sequence is also typed. Screenshots rot, and the letters in them cannot be copied cleanly.

Enzyme and PAM

Name the protein in the tube, not the family. SpCas9, SaCas9, a named engineered variant, and Cas12a are different orders. Next to the name, write the motif you required and which side of the protospacer it occupied. The motif is not part of the spacer. Recording it stops the next person from "correcting" your oligo by gluing NGG onto the end.

If the protein was a nickase or a dead variant, write that. A repeat that substitutes cutting Cas9 will not repeat a binding experiment. Motif differences are compared in PAM sequences and why the genome context matters.

Assembly, coordinates, and strand

A guide without a genome build is a sequence floating over several possible targets. Record the assembly or the accession, the chromosome or contig, the coordinates of the protospacer, and the strand. Ensembl and GenBank display names you can copy. "Human, exon 2" is not a coordinate. Two exons get numbered differently when transcripts differ. If you chose a transcript, record the transcript identifier.

Strand mistakes are the quiet failure. The spacer matches the target site in the orientation the enzyme needs. The genomic strand field tells a later reader which browser track was used. If you designed in a tool such as CRISPOR, save the enzyme setting and the genome setting with the output, not only the score. A high score on the wrong assembly is a confident miss.

When the parent line differs from the assembly, note the parental bases you actually checked, especially the motif. A repeat aimed at the reference will miss a strain that lost the PAM.

How it was assembled

Say whether the guide was a synthetic RNA, a synthetic DNA oligo cloned into a backbone, or a purchased plasmid. For a plasmid, record the map identifier, the promoter that drives the guide, the antibiotic marker, and where the spacer sits relative to the scaffold. A map that lives only as "final_v7" on one computer is not an identifier. Deposit the map where the group can open it, and write that location in the notebook.

If overhangs were added for cloning, write the spacer and the overhangs as separate fields. The classic repeat error is to reorder the whole oligo, overhangs included, as if it were the RNA spacer. The extra bases then become part of the match, or they prevent transcription from starting where you think.

Record the lot or synthesis batch that entered the cells. A later lot can differ in purity or in a silent resynthesis error. If you miniprepped the plasmid, record the prep date and who sequence-verified the spacer in that prep. A map is the design. The prep is the molecule. Both belong in the record. Verification sequencing habits for inserts are the same spirit as plasmid cloning from insert to colony.

Modifications

Chemical changes to a synthetic RNA, such as end protection used to slow degradation, change the molecule you ordered. Write them, or write none. Phosphorothioate linkages or terminal modifications that you did not mean to include will appear if someone accepts a supplier default. Defaults change. The notebook should not.

The same line should say the delivery class used with that lot: plasmid transfection, electroporation of RNA, or a ribonucleoprotein mixed from a stated protein lot and a stated RNA lot. The protein lot is part of the repeat. An RNA record without the protein it was complexed with is only half of an RNP experiment.

Failures that look like biology

A repeat that will not cut is often a record problem. The new oligo is the reverse complement. The PAM was included. The backbone promoter was for a different species' polymerase. The lot in the box is an older non-targeting control that shared a rack. None of these show up as a dramatic gel of the genomic locus. They show up as a wild-type sequence and a confident story about chromatin.

Branch the record before you branch the biology. Read the spacer aloud against the original design file. Check the enzyme name against the protein lot. Check that the coordinates still exist on the assembly you named. If the record cannot answer, the experiment cannot be repeated, and a new phenotype should not be compared with the old one.

Shared nicknames fail across buildings. Two students can both have a "guide 3" against different genes. The box label can stay short if the page it points to is unique. The page cannot stay short.

Humidity, heat, and labels

In a humid laboratory, adhesive labels slide off cold tubes when the tube warms. Write the identifier on the tube itself and in the book. RNA lots stored wet, without a dated resuspension note, cannot be assumed to match the dry pellet you ordered. Record the resuspension date and the buffer class. Follow the stability statement that came with the synthesis. A power cut that thaws a shared box is a reason to retire unmarked tubes, not a reason to guess which guide survived.

An enquiry to reorder should carry the fields, not the nickname. Put the enzyme, the spacer, the modifications, and the backbone on the quote request. Guide and nuclease classes are browsed in the molecular biology catalogue. If the repeat also needs the locus sequenced to show it matches the last experiment, use the CRISPR validation sequencing enquiry reference as a discussion prompt. It is an independent method reference. Ask whether a quotation is possible. The questions worth answering before that note goes out are in questions before an editing-validation enquiry. The molecular biology pathway is the research context.

A complete guide record is still only a research note. It does not approve the experiment, and it does not turn a repeated knockout into a diagnostic test. Biosafety for the repeat is the same institutional decision as for the first attempt, updated if the host or the vector changed.