Frequently asked questions
Getting Started
Do I need to create an account to use SciKeep?
No. Every analysis tool — qPCR ΔΔCt, the statistics engine, sequence tools, imaging, the protein and buffer calculators, the oligo calculator — work as a Guest with no account. An account saves your text records (notebook, protocols, inventories) to your profile so they survive a cleared cache; the $49/month Lab plan adds shared lab infrastructure for a whole group. No analysis tool is behind either.
Does SciKeep work offline?
Yes, for pages you have opened before. SciKeep installs a service worker that caches the application and the cell-segmentation models on your device, so any tool you have already visited opens and runs with no network at all — useful on a microscope workstation that is off the network by policy. Two honest limits: a page you have never opened on that device has nothing cached to fall back on, and the AI writing tools (Methods Generator, Protocol Generator, Peer Review Audit) contact a model provider and always need a connection. Every calculation itself runs in your browser either way, and your images and raw data never leave the device.
What costs money?
Every analysis tool is free forever — qPCR, statistics, imaging, IC50, the manuscript audit and the rest — with no account, no card and no trial clock. The only paid plan is Lab at $49/month, which adds shared infrastructure for the whole group: a shared -80°C freezer map, antibody and cell-line registries, a team notebook with a hash-chained audit trail, batch image analysis and reagent ordering. It is priced per lab rather than per seat, and there is a 14-day trial with email only, no credit card.
Can multiple people in my lab share a subscription?
The Lab plan ($49/month, or $39/month billed annually) is priced per lab rather than per seat — every member of the group is included, with shared access to the Part 11-ready Electronic Lab Notebook, -80°C Freezer Locator, Reagent Ordering Tracker, and .skvault disaster recovery. For departmental licenses (10+ seats), contact labs@scikeep.com.
Security & 21 CFR Part 11
How does the 21 CFR Part 11 SHA-256 cryptographic ledger work?
Every experiment creation, edit, sign-off, or deletion generates a SHA-256 hash block incorporating the previous block's hash, timestamp, user email, and action payload. This creates a tamper-evident chain: a record cannot be backdated or altered without the chain failing validation. Tamper-EVIDENT is not tamper-proof — the ledger lives in your own browser storage, so someone with the developer console can rewrite the whole chain, and clearing site data removes it. What the chain gives you is detection, not prevention. Note also that tamper-evidence is one Part 11 control among several — SciKeep is not itself a validated Part 11 system, and compliance is a property of your institution's validated installation and procedures.
Are digital electronic signatures legally binding for patent defensibility?
Electronic signatures in SciKeep are architected around FDA 21 CFR Part 11 requirements for closed workflows, including cryptographic SHA-256 seal chaining, explicit declaration of signer role (Author, Reviewer, PI), and tamper-evident audit logs. For enterprise institutional validation and legal audit certifications, contact labs@scikeep.com.
What is the .skvault disaster recovery archive?
The .skvault archive is a 1-click encrypted snapshot of your entire laboratory database: all ELN experiment logs, 21 CFR Part 11 audit trails, -80°C freezer rack maps, antibody banks, and calibration curves. You can download and restore this file on any machine with 1 click.
Where is my research data stored? Does SciKeep see my sequences?
As a guest, everything stays in your browser's local storage — nothing reaches us and we cannot access it, but a cleared cache loses it permanently. If you create an account, your text records (notebook entries and inventories) are also saved to your private account so they survive a wipe; only you can read them. Images and raw data never sync in either mode — they are stripped before upload and the database rejects anything image-sized.
Scientific Accuracy
How are the algorithms validated? Can I cite them in publications?
The calculation tools implement published, peer-reviewed methods: SantaLucia (1998) unified nearest-neighbor thermodynamics for DNA Tm, Pace et al. (1995) for protein extinction, Livak & Schmittgen (2001) for qPCR 2^-ddCt, grid-search plus coordinate-descent least squares for 4PL/5PL IC50, Lord et al. (2020) for SuperPlots replicate grouping, and Mantel-Cox log-rank test for Kaplan-Meier survival curves. The engines are covered by an automated test suite of several hundred tests that checks results against published worked examples and reference values where those exist, and each tool carries an evidence grade on the validation page (/validation) so you can see how strongly it has been checked.
What are SuperPlots and why are they recommended for microscopy quantification?
Standard cell biology plots pool thousands of individual cell observations (n=1,500) into a single statistical test, falsely inflating statistical significance through pseudoreplication. SuperPlots (Lord et al. 2020) preserve biological replicate identities (N=3) by color-coding individual cell points and computing true cluster-level p-values using the means of the independent biological experiments.
How does ROUT outlier detection work in 96-well plate assays?
The ROUT method (Motulsky & Brown 2006) uses robust non-linear regression combined with False Discovery Rate (FDR Q=1%) thresholding based on the Median Absolute Deviation (MAD) of residuals. It objectively detects and excludes dispensing bubbles and edge evaporation artifacts without biasing the fitted IC50/EC50 curve.
How does the Biopharm mAb formulation suite calculate viscosity and stability?
Buffer capacity is calculated using the exact Van Slyke equation (β = 2.303·C·Ka·[H+] / (Ka + [H+])²). High-concentration antibody viscosity (up to 250 mg/mL) is calculated via the Mooney / Ross-Minton equation accounting for electroviscous crowding and L-Arginine self-association suppression. Electrostatic screening length is determined via Debye-Hückel theory (κ⁻¹ = 0.304 / √I nm).
How does Multiple Sequence Alignment (MSA) work?
SciKeep implements progressive global Needleman-Wunsch pairwise alignment with affine gap penalty matrices and UPGMA guide tree construction. It computes full Clustal conservation consensus symbols (*, :, .) with color-coded amino acid chemistry and 1-click FASTA export.
How does 4PL vs 5PL asymmetric logistic curve fitting work?
The 4PL model fits symmetric sigmoidal curves (y = Bottom + (Top - Bottom) / (1 + (x/IC50)^Hill)). For immunoassays (ELISA, AlphaLISA) and monoclonal antibody binding where upper and lower asymptotes exhibit unequal curvature, the 5PL engine introduces an asymmetry parameter S: y = Bottom + (Top - Bottom) / (1 + (x/IC50)^Hill)^S.
Plans, Billing & University POs
Can our university pay via Purchase Order (PO) or Grant code?
Card payment through Stripe is available now and produces an itemised receipt you can use for grant reimbursement, including on a departmental or grant purchasing card. For a purchase order, email billing@scikeep.com with your institution's requirements and we will tell you plainly whether we can meet them — SciKeep is an early-stage vendor, so we cannot yet meet every supplier-onboarding requirement, and wire and ACH bank transfer are not offered at present.
How do I download receipts for university grant reimbursement?
Signed-in users can open the "Researcher Account & Workspace" modal, click "Subscription & Billing", and download itemized official receipts with 1 click for any billing period.
Is there a discount for annual billing?
Yes. The Lab plan is $49/month, or $468/year billed annually ($39/month equivalent) — a 20% discount. It is priced per lab, not per seat.
How do I cancel? What happens to my data?
You can cancel anytime from your billing settings. Your data remains safe and accessible in your local browser. You can also export your full laboratory state via .skvault before canceling.