Figure S1. Agarose gel stained with ethidium bromide for CSP-CIPer detection limit of HPV -56 in presence of human genomic DNA.
The plasmid concentration was varied, ranging from 10 ng to 100 fg per reaction at a constant background of 200 ng non-HPV-contaminated human genomic DNA. The minimum detectable amount of HPV observed was 1 pg, and the 100 fg mixture showed no significant amplification. The upper triangular graph represents a visual interpretation of CIPer amplicon intensity as a function of plasmid concentration, and the lower graph the background intensity as a function of plasmid concentration.

Figure S2. Agarose gel stained with ethidium bromide for fragment size determination and pyrograms for HPV genotypes -16 and -18.
Amplicon sizes appear in the expected size of 185 base pairs for HPV-16 and 188 base pairs for HPV-18. The negative controls for the CIPer reaction (“CIPer-”) and PCR (“PCR-”) show no significant amplification. Pyrograms derived from multiple sequencing primers MSP-16 and MSP-18 (Table S1) validate the expected sequence.

Figure S3. Genomic regions GP5+/6+ aligned for 40 genotypes detected by either CSP-CIPer or GP5+/6+ PCR (Table 1).
Alignment was performed and displayed with ClustalX (http://www.biodirectory.com/biowiki/ClustalX). The targeted region flanking the two primers varies between 90-100 base pairs depending on genotype characteristics, and is represented in the figure with three Ns.

Figure S4. Detecting multiple HPV co-infections.
In order to compare the discriminative power of the CIPer vs. PCR detection, artificial mixtures of the eight plasmids HPV-16, -18, -33, -35, -39, -45, -58 and -59 were constructed to “mimic” real-case multiple co-infections. The CIPer method could detect all eight genotypes present in the same sample, while PCR managed with the four genotypes HPV-16, -18, -33 and -45. As seen in the figure the Pyrogram intensities vary for different genotypes, indicating a lower discriminative preference for certain types. Among the four PCR detected genotypes, HPV-16 and -33 showed very weak diagrams and were barely detectable in the presence of the preferred genotypes -18 and -45. CIPer detection suffered to a lesser degree of such preferred selectivity, as all types were clearly distinguishable. We believe that differences in target amplification strategy account for this improvement, i.e. PCR is based on target dependent amplification, while the CIPer is based on target dependent probe circularization, followed by non-target dependent universal amplification.

Figure S5. Digital amplification strategies prior to probes validation.
A) Emulsion based PCR; adapters are optional since CIPers already contain universal segments flanking the target of interest. The technology involves the inclusion of DNA and a primed magnetic bead in mineral oil (an emulsion), which allows for single molecule amplifications. Suitable upcoming methods for downstream sequence validation include the 454 Life Sciences platform (http://www.454.com), the Solexa platform (http://www.illumina.com), polony sequencing (http://www.agentcourt.com), Helicos Biosciences (http://www.helicosbio.com), or use of a resequencing array (GeneChip CustomSeq, www.affymetrix.com). B) Rolling circle amplification (RCA) with single-molecule detection (SMD). Digital quantification combines RCA and SMD in form of fluorescent-labeled target specific oligonucleotides. The amplified CIPers can be quantified using microfluidic analysis and visualized with a microscope for ultimate levels of quantification. The number of available fluorescent labels limits the degree of multiplexing.

Figure S6. “Multiplex multiplexing padlocks” (MMP) construct assembly for CIPer probes production.
The extended CIPer carries a third universal segment and flanks the barcode together with the other universal segments. The probe is divided and synthesized in three constructs, with the middle one containing the barcode. Through a two-way ligation scheme the three fragments can be joined into a full-length probe. Following the grid-like assembly strategy displayed in the figure, one-genotype detection pair of AS/ES constructs is combined with multiple unique barcodes. The unique barcode now carries two dimensions of information, i.e. genotype and patient ID. Post-reaction CIPers can be combined into multiple patient pools and all screened simultaneously, essentially lowering costs involved with downstream validations, both for hybridization-based techniques and sequencing procedures.

Table S1. Oligonucleotides used in the present study.
Bold italic marked sequences in the anchor sites of CIPers -gyrA and +parC were added to the originally described primers to obtain higher annealing temperature values for the anchor site. The number in parenthesis following the multiple sequencing primers (MSPs) denotes which primer pool the MSP belongs to.