spec/technologies/cel_seq.json

{
  "schema_version": "seqcolyte.spec.v1",
  "spec_id": "cel_seq",
  "assay": "CEL-Seq",
  "chemistry_version": "",
  "platform": "illumina",
  "platform_params": {
    "read_type": "short"
  },
  "source_docs": [
    {
      "doc_id": "CEL-Seq_protocol.pdf",
      "title": "CEL-Seq_protocol.pdf",
      "url": null,
      "path": "/Users/seqmachines/playground/protocols-test/protocols/cel_seq/CEL-Seq_protocol.pdf",
      "retrieved_date": null
    },
    {
      "doc_id": "CEL-Seq_paper.pdf",
      "title": "CEL-Seq_paper.pdf",
      "url": null,
      "path": "/Users/seqmachines/playground/protocols-test/protocols/cel_seq/CEL-Seq_paper.pdf",
      "retrieved_date": null
    }
  ],
  "oligos": [
    {
      "oligo_id": "celseq_rt_primer",
      "name": "CEL-Seq barcoded RT primer",
      "aliases": [],
      "role": "Reverse-transcription / capture primer carrying anchored poly(dT), cell barcode, 5' Illumina small-RNA adapter and T7 promoter",
      "kind": "assembled",
      "sequence": "CGATTGAGGCCGGTAATACGACTCACTATAGGGGTTCAGAGTTCTACAGTCCGACGATC[CELL_BARCODE:8]TTTTTTTTTTTTTTTTTTTTTTTTV",
      "direction": "5_to_3",
      "components": [
        {
          "name": "5' leader",
          "sequence": "CGATTGAGGCCGG",
          "role": "5' linker/leader upstream of T7 promoter"
        },
        {
          "name": "T7 promoter",
          "sequence": "TAATACGACTCACTATAGGG",
          "role": "T7 RNA polymerase promoter (drives IVT), as in Baugh et al. 2001"
        },
        {
          "name": "Illumina 5' RNA adapter (RA5)",
          "sequence": "GTTCAGAGTTCTACAGTCCGACGATC",
          "role": "5' Illumina small-RNA sequencing adapter"
        },
        {
          "name": "Cell barcode",
          "sequence": "[CELL_BARCODE:8]",
          "role": "8-bp cell/sample barcode read at the start of read 1"
        },
        {
          "name": "Anchored poly(dT)",
          "sequence": "TTTTTTTTTTTTTTTTTTTTTTTTV",
          "role": "24-nt poly(dT) with 3' V anchor for polyA mRNA capture"
        }
      ],
      "provenance": "document",
      "derivation": null,
      "sequence_source": "llm_extracted_from_docs",
      "evidence": [
        {
          "source_doc": "protocol_docs",
          "locator": "oligo / final library",
          "method": "claude_llm_extraction"
        }
      ],
      "notes": "Working conc. 25 ng/ul, desalted. Barcodes are 8 bp, designed in groups of four so the first five positions have equal base representation; each pair differs by >=2 nt. The 40 published barcodes (inserted at [CELL_BARCODE:8], 5'->3'): 1 CATCACGC, 2 GTCGTCGC, 3 ACGACCGC, 4 TGATGCGC, 5 CATCAATC, 6 GTCGTATC, 7 ACGACATC, 8 TGATGATC, 9 CATCATCC, 10 GTCGTTCC, 11 ACGACTCC, 12 TGATGTCC, 13 CATCAGAA, 14 GTCGTGAA, 15 ACGACGAA, 16 TGATGGAA, 17 TCACACGC, 18 CGTGTCGC, 19 GACACCGC, 20 ATGTGCGC, 21 TCACATCA, 22 CGTGTTCA, 23 GACACTCA, 24 ATGTGTCA, 25 TCACAGAG, 26 CGTGTGAG, 27 GACACGAG, 28 ATGTGGAG, 29 CTAACCGC, 30 GCTTGCGC, 31 AGCCACGC, 32 TAGGTCGC, 33 CTAACTCA, 34 GCTTGTCA, 35 AGCCATCA, 36 TAGGTTCA, 37 CTAACGAG, 38 GCTTGGAG, 39 AGCCAGAG, 40 TAGGTGAG."
    },
    {
      "oligo_id": "illumina_5p_rna_adapter_ra5",
      "name": "Illumina 5' RNA adapter (RA5)",
      "aliases": [],
      "role": "5' Illumina small-RNA adapter; supplied within the CEL-Seq RT primer",
      "kind": "single",
      "sequence": "GTTCAGAGTTCTACAGTCCGACGATC",
      "direction": "5_to_3",
      "components": [],
      "provenance": "document",
      "derivation": null,
      "sequence_source": "llm_extracted_from_docs",
      "evidence": [
        {
          "source_doc": "protocol_docs",
          "locator": "oligo / final library",
          "method": "claude_llm_extraction"
        }
      ],
      "notes": "5' Illumina adapter 'as used in the Illumina small RNA kit'; this segment is embedded in the CEL-Seq RT primer, so its sequence is documented directly."
    },
    {
      "oligo_id": "illumina_3p_rna_adapter_ra3",
      "name": "Illumina 3' RNA adapter (RA3)",
      "aliases": [],
      "role": "3' Illumina small-RNA adapter ligated to fragmented aRNA",
      "kind": "single",
      "sequence": "TGGAATTCTCGGGTGCCAAGG",
      "direction": "5_to_3",
      "components": [],
      "provenance": "document",
      "derivation": null,
      "sequence_source": "llm_extracted_from_docs",
      "evidence": [
        {
          "source_doc": "protocol_docs",
          "locator": "oligo / final library",
          "method": "claude_llm_extraction"
        }
      ],
      "notes": "TruSeq Small RNA RA3 adapter; protocol uses it diluted 1:5. Sequence cited in the protocol as 'available from Illumina' and not printed there; shown is the standard Illumina TruSeq Small RNA RA3 sequence."
    },
    {
      "oligo_id": "rna_rt_primer_rtp",
      "name": "RNA RT Primer (RTP)",
      "aliases": [],
      "role": "Reverse-transcription primer complementary to the ligated 3' adapter (RA3)",
      "kind": "single",
      "sequence": "GCCTTGGCACCCGAGAATTCCA",
      "direction": "5_to_3",
      "components": [],
      "provenance": "document",
      "derivation": null,
      "sequence_source": "llm_extracted_from_docs",
      "evidence": [
        {
          "source_doc": "protocol_docs",
          "locator": "oligo / final library",
          "method": "claude_llm_extraction"
        }
      ],
      "notes": "From the Illumina TruSeq Small RNA kit; primes SuperScript II RT of the RA3-ligated RNA. Sequence not printed in the protocol ('available from Illumina'); shown is the standard TruSeq Small RNA RTP."
    },
    {
      "oligo_id": "rna_pcr_primer_rp1",
      "name": "RNA PCR Primer (RP1)",
      "aliases": [],
      "role": "Forward library PCR primer (adds P5 and completes the 5' adapter)",
      "kind": "assembled",
      "sequence": "AATGATACGGCGACCACCGAGATCTACACGTTCAGAGTTCTACAGTCCGA",
      "direction": "5_to_3",
      "components": [
        {
          "name": "P5",
          "sequence": "AATGATACGGCGACCACCGAGATCTACAC",
          "role": "Illumina P5 flow-cell adapter"
        },
        {
          "name": "Partial RA5",
          "sequence": "GTTCAGAGTTCTACAGTCCGA",
          "role": "5' small-RNA adapter overlap"
        }
      ],
      "provenance": "document",
      "derivation": null,
      "sequence_source": "llm_extracted_from_docs",
      "evidence": [
        {
          "source_doc": "protocol_docs",
          "locator": "oligo / final library",
          "method": "claude_llm_extraction"
        }
      ],
      "notes": "From the Illumina TruSeq Small RNA kit. Sequence not printed in the protocol ('available from Illumina'); shown is the standard TruSeq Small RNA RP1."
    },
    {
      "oligo_id": "indexed_rna_pcr_primer_rpix",
      "name": "Indexed RNA PCR Primer (RPIX)",
      "aliases": [],
      "role": "Reverse indexed library PCR primer (adds i7 sample index and P7)",
      "kind": "assembled",
      "sequence": "CAAGCAGAAGACGGCATACGAGAT[SAMPLE_INDEX:6]GTGACTGGAGTTCCTTGGCACCCGAGAATTCCA",
      "direction": "5_to_3",
      "components": [
        {
          "name": "P7",
          "sequence": "CAAGCAGAAGACGGCATACGAGAT",
          "role": "Illumina P7 flow-cell adapter"
        },
        {
          "name": "i7 sample index",
          "sequence": "[SAMPLE_INDEX:6]",
          "role": "6-nt Illumina library index (read as separate index read)"
        },
        {
          "name": "RA3/RTP overlap region",
          "sequence": "GTGACTGGAGTTCCTTGGCACCCGAGAATTCCA",
          "role": "3' adapter overlap, complementary to RA3"
        }
      ],
      "provenance": "document",
      "derivation": null,
      "sequence_source": "llm_extracted_from_docs",
      "evidence": [
        {
          "source_doc": "protocol_docs",
          "locator": "oligo / final library",
          "method": "claude_llm_extraction"
        }
      ],
      "notes": "RPIX = uniquely indexed reverse PCR primer from the Illumina TruSeq Small RNA kit; provides an additional per-library barcode so multiple IVT pools run together. Sequence not printed in the protocol ('available from Illumina'); shown is the standard TruSeq Small RNA RPI scaffold."
    },
    {
      "oligo_id": "illumina_p5_adapter",
      "name": "Illumina P5 adapter",
      "aliases": [],
      "role": "P5 flow-cell binding sequence (5' end of final library)",
      "kind": "single",
      "sequence": "AATGATACGGCGACCACCGAGATCTACAC",
      "direction": "5_to_3",
      "components": [],
      "provenance": "document",
      "derivation": null,
      "sequence_source": "llm_extracted_from_docs",
      "evidence": [
        {
          "source_doc": "protocol_docs",
          "locator": "oligo / final library",
          "method": "claude_llm_extraction"
        }
      ],
      "notes": "Introduced via RP1; canonical Illumina P5."
    },
    {
      "oligo_id": "illumina_p7_adapter",
      "name": "Illumina P7 adapter",
      "aliases": [],
      "role": "P7 flow-cell binding sequence (3' end of final library)",
      "kind": "single",
      "sequence": "CAAGCAGAAGACGGCATACGAGAT",
      "direction": "5_to_3",
      "components": [],
      "provenance": "document",
      "derivation": null,
      "sequence_source": "llm_extracted_from_docs",
      "evidence": [
        {
          "source_doc": "protocol_docs",
          "locator": "oligo / final library",
          "method": "claude_llm_extraction"
        }
      ],
      "notes": "Introduced via RPIX; canonical Illumina P7. Appears in the final library as its reverse complement ATCTCGTATGCCGTCTTCTGCTTG."
    }
  ],
  "final_library": {
    "source_label": "CEL-Seq final Illumina (TruSeq Small RNA) library",
    "annotated_library_sequence": "AATGATACGGCGACCACCGAGATCTACACGTTCAGAGTTCTACAGTCCGACGATC[CELL_BARCODE:8]TTTTTTTTTTTTTTTTTTTTTTTT[CDNA]TGGAATTCTCGGGTGCCAAGGAACTCCAGTCAC[SAMPLE_INDEX:6]ATCTCGTATGCCGTCTTCTGCTTG",
    "library_sequence": "AATGATACGGCGACCACCGAGATCTACACGTTCAGAGTTCTACAGTCCGACGATCNNNNNNNNTTTTTTTTTTTTTTTTTTTTTTTTXXXXXXXXXXTGGAATTCTCGGGTGCCAAGGAACTCCAGTCACNNNNNNATCTCGTATGCCGTCTTCTGCTTG",
    "strands": [
      {
        "direction": "5_to_3",
        "source_html": "AATGATACGGCGACCACCGAGATCTACACGTTCAGAGTTCTACAGTCCGACGATCNNNNNNNNTTTTTTTTTTTTTTTTTTTTTTTT[CDNA]TGGAATTCTCGGGTGCCAAGGAACTCCAGTCACNNNNNNATCTCGTATGCCGTCTTCTGCTTG",
        "source_sequence": "AATGATACGGCGACCACCGAGATCTACACGTTCAGAGTTCTACAGTCCGACGATCNNNNNNNNTTTTTTTTTTTTTTTTTTTTTTTT[CDNA]TGGAATTCTCGGGTGCCAAGGAACTCCAGTCACNNNNNNATCTCGTATGCCGTCTTCTGCTTG"
      }
    ],
    "annotation_lines": [
      "AATGATACGGCGACCACCGAGATCTACAC = P5",
      "GTTCAGAGTTCTACAGTCCGACGATC = Illumina 5' RNA adapter (RA5) / read 1 region",
      "[CELL_BARCODE:8] = Cell barcode (CEL-Seq RT-primer barcode)",
      "TTTTTTTTTTTTTTTTTTTTTTTT = poly(dT) (mRNA polyA-derived)",
      "[CDNA] = cDNA insert (3'-most fragment of the transcript)",
      "TGGAATTCTCGGGTGCCAAGG = Illumina 3' RNA adapter (RA3)",
      "AACTCCAGTCAC = TruSeq Small RNA read 2 / RP linker",
      "[SAMPLE_INDEX:6] = i7 sample index (RPIX)",
      "ATCTCGTATGCCGTCTTCTGCTTG = reverse complement of P7"
    ],
    "evidence": [
      {
        "source_doc": "protocol_docs",
        "locator": "CEL-Seq final Illumina (TruSeq Small RNA) library",
        "method": "claude_llm_extraction"
      }
    ]
  },
  "read_structure": {
    "reads": [
      {
        "read": "R1",
        "primer": "Illumina small-RNA Read 1 sequencing primer (RA5)",
        "template": "bottom",
        "cycles": 15,
        "segments": [
          {
            "name": "Cell barcode",
            "type": "barcode",
            "order": 0,
            "scored": true,
            "provenance": null,
            "whitelist_ref": null,
            "constant_ref": null,
            "notes": null,
            "length": 8
          },
          {
            "name": "poly(dT)",
            "type": "polyA",
            "order": 1,
            "scored": false,
            "provenance": null,
            "whitelist_ref": null,
            "constant_ref": null,
            "notes": null,
            "length": 7
          }
        ]
      },
      {
        "read": "I1",
        "primer": "Illumina index sequencing primer",
        "template": "top",
        "cycles": 6,
        "segments": [
          {
            "name": "i7 sample index",
            "type": "index",
            "order": 0,
            "scored": false,
            "provenance": null,
            "whitelist_ref": null,
            "constant_ref": null,
            "notes": null,
            "length": 6
          }
        ]
      },
      {
        "read": "R2",
        "primer": "Illumina small-RNA Read 2 sequencing primer (RA3/RP)",
        "template": "top",
        "cycles": 50,
        "segments": [
          {
            "name": "cDNA insert",
            "type": "insert",
            "order": 0,
            "scored": true,
            "provenance": null,
            "whitelist_ref": null,
            "constant_ref": null,
            "notes": null
          }
        ]
      }
    ]
  },
  "library_generation": [
    {
      "step": 1,
      "title": "Single-cell lysis, primer annealing & first-strand synthesis",
      "summary": "The barcoded CEL-Seq primer captures the cell's polyA mRNA and ArrayScript reverse transcriptase copies it into first-strand cDNA.",
      "note": "Each frozen single cell receives one uniquely barcoded CEL-Seq RT primer (+ERCC spike-in), is heated to open the cell and anneal the anchored poly(dT), then reverse transcribed at 42C (Ambion MessageAmp II ArrayScript).",
      "product": "         T7 promoter                RA5 (5' adapter)          BC        poly(dT)\n5'- CGATTGAGGCCGGTAATACGACTCACTATAGGGGTTCAGAGTTCTACAGTCCGACGATC[CELL_BARCODE:8]TTTTTTTTTTTTTTTTTTTTTTTTV[CDNA]------->\n                                                                                      3'- ...AAAAAAAAAA[mRNA] -5'"
    },
    {
      "step": 2,
      "title": "Second-strand synthesis",
      "summary": "RNaseH + DNA Pol I convert the cDNA:mRNA hybrid into double-stranded cDNA carrying a functional T7 promoter.",
      "note": "Standard second-strand reaction at 16C yields dsDNA; the T7 promoter is now double-stranded and transcription-competent.",
      "product": "5'- CGATTGAGGCCGGTAATACGACTCACTATAGGGGTTCAGAGTTCTACAGTCCGACGATC[CELL_BARCODE:8]TTTTTTTTTTTTTTTTTTTTTTTTV[CDNA] -3'\n3'- GCTAACTCCGGCCATTATGCTGAGTGATATCCCCAAGTCTCAAGATGTCAGGCTGCTAG[CELL_BARCODE:8]AAAAAAAAAAAAAAAAAAAAAAAAB[CDNA] -5'"
    },
    {
      "step": 3,
      "title": "Pooling + T7 in vitro transcription (linear amplification)",
      "summary": "Barcoded double-stranded cDNAs from many cells are pooled and linearly amplified in one round of T7 IVT into antisense amplified RNA (aRNA).",
      "note": "Pooling reaches the ~400 pg IVT input threshold; a single 13 hr IVT reaction makes many antisense RNA copies per molecule, each retaining the barcode and 5' adapter.",
      "product": "aRNA (antisense, 5'->3'; hundreds of copies):\n5'- GGGUUCAGAGUUCUACAGUCCGACGAUC[CELL_BARCODE:8]UUUUUUUUUUUUUUUUUUUUUUUU[antisense-cDNA] -3'"
    },
    {
      "step": 4,
      "title": "RNA fragmentation",
      "summary": "aRNA is chemically fragmented to a sequencing-appropriate size, peaking around 500 nt.",
      "note": "Fragmentation (Tris-acetate/KOAc/MgOAc, 94C, 3 min) breaks aRNA; only the 5'-most fragment still carries the barcode + RA5 adapter.",
      "product": "5'- GGGUUCAGAGUUCUACAGUCCGACGAUC[CELL_BARCODE:8]UUUUUUUUUUUU[frag] -3'   (barcode-bearing fragment)\n5'- [internal aRNA fragments, no adapter] -3'                            (not retained in final library)"
    },
    {
      "step": 5,
      "title": "End repair (Antarctic phosphatase + PNK)",
      "summary": "Fragment ends are dephosphorylated then re-phosphorylated to give 5'-monophosphate / 3'-OH ends compatible with adapter ligation.",
      "note": "Antarctic phosphatase removes 3' phosphates/cyclic phosphates; T4 PNK adds a 5' phosphate, priming the RNA for T4 RNA ligase 2 (truncated).",
      "product": "5'-P- GGGUUCAGAGUUCUACAGUCCGACGAUC[CELL_BARCODE:8]UUUUUUUUUUUU[frag] -OH-3'"
    },
    {
      "step": 6,
      "title": "3' adapter (RA3) ligation",
      "summary": "The Illumina 3' small-RNA adapter (RA3) is ligated onto the 3' end of the aRNA fragments.",
      "note": "T4 RNA Ligase 2 (truncated) attaches RA3 (diluted 1:5) to 3'-OH ends; only fragments with both the 5' (RA5, from the primer) and 3' (RA3) adapters become amplifiable.",
      "product": "5'- GGGUUCAGAGUUCUACAGUCCGACGAUC[CELL_BARCODE:8]UUUUUUUUUUUU[frag]TGGAATTCTCGGGTGCCAAGG -3'"
    },
    {
      "step": 7,
      "title": "Reverse transcription (RTP primer)",
      "summary": "The RNA RT primer (RTP) anneals to the ligated RA3 adapter and SuperScript II copies the adapter-flanked RNA into cDNA.",
      "note": "RTP is complementary to RA3; RT at 50C generates a cDNA copy that is now the PCR template.",
      "product": "RNA:  5'- ...GTTCAGAGTTCTACAGTCCGACGATC[CELL_BARCODE:8](T)24[frag]TGGAATTCTCGGGTGCCAAGG -3'\nRTP:                                                              3'- ACCTTAAGAGCCCACGGTTCCG -5'\ncDNA: 3'- ...CAAGTCTCAAGATGTCAGGCTGCTAG[CELL_BARCODE:8](A)24[frag]ACCTTAAGAGCCCACGGTTCCG -5'"
    },
    {
      "step": 8,
      "title": "Library PCR (RP1 + indexed RPIX) & bead cleanup",
      "summary": "12 cycles of PCR with RP1 (P5) and an indexed RPIX (i7 + P7) complete both flow-cell adapters and add the library index, producing the final dsDNA library.",
      "note": "RP1 adds P5/completes RA5; RPIX adds the 6-nt i7 index and P7. Two AMPure XP cleanups yield a ~300-400 bp library.",
      "product": "5'- AATGATACGGCGACCACCGAGATCTACACGTTCAGAGTTCTACAGTCCGACGATC[CELL_BARCODE:8]TTTTTTTTTTTTTTTTTTTTTTTT[CDNA]TGGAATTCTCGGGTGCCAAGGAACTCCAGTCAC[SAMPLE_INDEX:6]ATCTCGTATGCCGTCTTCTGCTTG -3'"
    }
  ],
  "library_sequencing": [
    {
      "read": "Read 1 (cell barcode + poly(T))",
      "primer": "Illumina small-RNA Read 1 sequencing primer (sequence = RA5; anneals to the bottom strand across the 5' RNA adapter)",
      "template": "bottom",
      "cycles": 15,
      "note": "Reads the 8-bp cell barcode immediately followed by the anchored poly(T) stretch (>=15 bases were read for read 1); >95% of reads carry the barcode precisely at the start of read 1, invariably followed by poly(T).",
      "diagram": "Read 1 primer (= RA5) anneals to bottom strand; reads cell barcode then poly(T):\n5'-...GTTCAGAGTTCTACAGTCCGACGATC----------------->            (read 1, 5'->3')\n5'-...GTTCAGAGTTCTACAGTCCGACGATCNNNNNNNNTTTTTTT...[cDNA]-3'   (top)\n3'-...CAAGTCTCAAGATGTCAGGCTGCTAGNNNNNNNNAAAAAAA...[cDNA]-5'   (bottom = template)\n                              ^cell BC(8)^ ^poly(T)^"
    },
    {
      "read": "Index 1 (i7 sample index)",
      "primer": "Illumina i7 index sequencing primer (anneals across the RP/RA3 linker on the top strand, 3' end abutting the index)",
      "template": "bottom",
      "cycles": 6,
      "note": "Reads the 6-nt library-level index carried on the indexed RNA PCR primer (RPIX); used to demultiplex multiple IVT pools sequenced on the same lane. Read is only taken when multiplexing IVT pools.",
      "diagram": "i7 index primer anneals over RP linker; reads the 6-nt sample index:\n5'-...TGGAATTCTCGGGTGCCAAGGAACTCCAGTCAC------>                        (index read, 5'->3')\n5'-...TGGAATTCTCGGGTGCCAAGGAACTCCAGTCACNNNNNNATCTCGTATGCCGTCTTCTGCTTG-3'  (top)\n3'-...ACCTTAAGAGCCCACGGTTCCTTGAGGTCAGTGNNNNNNTAGAGCATACGGCAGAAGACGAAC-5'  (bottom = template)\n                                        ^i7 index(6)^"
    },
    {
      "read": "Read 2 (cDNA / mRNA transcript)",
      "primer": "Illumina small-RNA Read 2 sequencing primer (sequence complementary to RA3; anneals to the top strand across the 3' RNA adapter)",
      "template": "top",
      "cycles": 50,
      "note": "Reads 50 bases into the 3'-most cDNA fragment, identifying the transcript strand-specifically (>97% of exonic reads sense-oriented; reads map to transcript 3' ends).",
      "diagram": "Read 2 primer anneals over RA3 on top strand; reads into 3'-most cDNA fragment (antisense):\n                                 3'-ACCTTAAGAGCCCACGGTTCC-5'                 (Read 2 primer)\n                                 <------(read 2)\n5'-...TTTTTTT[.....cDNA insert.....]TGGAATTCTCGGGTGCCAAGGAACTCCAGTCAC...-3'  (top = template)\n3'-...AAAAAAA[.....cDNA insert.....]ACCTTAAGAGCCCACGGTTCCTTGAGGTCAGTG...-5'  (bottom)"
    }
  ],
  "whitelists": {},
  "build": {
    "builder_version": "llm-generic-1.0",
    "deterministic": false,
    "source_html_sha256": null,
    "extraction_method": "claude_llm_generic",
    "model": "claude-opus-4-8"
  },
  "title": "CEL-Seq",
  "description": "CEL-Seq (Cell Expression by Linear amplification and Sequencing) is a multiplexed single-cell RNA-seq method that barcodes each cell's mRNA during reverse transcription, pools the samples, and linearly amplifies them with a single round of T7 in vitro transcription (IVT) instead of PCR. Each cell's RT primer carries an anchored poly(dT), an 8-bp cell barcode, the Illumina 5' small-RNA adapter, and a T7 promoter. After IVT the amplified antisense RNA is fragmented and built into an Illumina directional (TruSeq Small RNA) library. Paired-end sequencing reads the cell barcode followed by a poly(T) stretch in read 1 and the 3'-most mRNA fragment in read 2, yielding strand-specific, 3'-end digital gene-expression counts.",
  "reference": {
    "kind": "protocol_doc",
    "label": "cel seq",
    "path": null,
    "url": null,
    "doi": null
  },
  "publication": {
    "year": 2012,
    "original_publication": {
      "title": "CEL-Seq: Single-Cell RNA-Seq by Multiplexed Linear Amplification",
      "journal": "Cell Reports",
      "doi": "10.1016/j.celrep.2012.08.003",
      "url": "http://dx.doi.org/10.1016/j.celrep.2012.08.003"
    },
    "authors": [
      {
        "name": "Tamar Hashimshony",
        "affiliation": "Department of Biology, Technion-Israel Institute of Technology, Haifa 32000, Israel"
      },
      {
        "name": "Florian Wagner",
        "affiliation": "Department of Biology, Technion-Israel Institute of Technology, Haifa 32000, Israel"
      },
      {
        "name": "Noa Sher",
        "affiliation": "Department of Biology, Technion-Israel Institute of Technology, Haifa 32000, Israel"
      },
      {
        "name": "Itai Yanai",
        "corresponding": true,
        "email": "yanai@technion.ac.il",
        "affiliation": "Department of Biology, Technion-Israel Institute of Technology, Haifa 32000, Israel"
      }
    ],
    "throughput": {
      "summary": "Amplification of up to ~50 cells per person per day; ~10 IVT pools (~500 cells) converge into a single library preparation. The 8-bp barcode design makes the number of uniquely barcoded samples per IVT essentially unlimited.",
      "cells": "~500 single cells per library prep",
      "rna": "IVT requires ~400 pg total RNA input; RT works down to ~5 pg (single mammalian cell). 10 pg C. elegans RNA ~ 280,000 mRNA molecules.",
      "dna": "Final library ~1 ng/ul, size peak ~300-400 bp"
    },
    "statistical_model": "Absolute molecule counts obtained by least-squares linear regression on ERCC spike-in expression levels; per-gene expression reported as transcripts per million (tpm), with differential expression tested by FDR-corrected t tests.",
    "other": [
      {
        "label": "NCBI SRA accession",
        "value": "SRP014672"
      },
      {
        "label": "Amplification",
        "value": "Single round of linear T7 in vitro transcription (IVT), not PCR"
      },
      {
        "label": "Strand specificity",
        "value": ">98% of exonic reads from the sense strand"
      },
      {
        "label": "Barcoding efficiency",
        "value": ">96% of reads contain a barcode (95.5% at start of read 1)"
      },
      {
        "label": "rRNA contamination",
        "value": "<2% of reads (polyA-specific)"
      },
      {
        "label": "Kits",
        "value": "Ambion MessageAmp II aRNA (AM1751) + Illumina TruSeq Small RNA sample prep (RS-200-0012)"
      },
      {
        "label": "Sequencer",
        "value": "Illumina HiSeq2000, paired-end"
      }
    ]
  },
  "modality": "RNA",
  "method_type": "manual/tube",
  "data_processing": {
    "summary": "A custom multistep pipeline run in the Galaxy framework: paired-end reads are trimmed/filtered, demultiplexed by the first 8 bases of read 1 (the cell barcode), read 2 is mapped to the reference genome, reads are counted per gene and normalized to transcripts per million (tpm). Because CEL-Seq retains a single 3'-most fragment per transcript, counts require no gene-length normalization; absolute mRNA copy numbers are then obtained by least-squares linear regression on the ERCC spike-in values.",
    "stages": [
      {
        "id": "preprocess",
        "label": "Read preprocessing"
      },
      {
        "id": "demultiplex",
        "label": "Demultiplexing"
      },
      {
        "id": "align",
        "label": "Alignment"
      },
      {
        "id": "quantify",
        "label": "Quantification"
      },
      {
        "id": "normalize",
        "label": "Normalization & absolute counts"
      }
    ],
    "nodes": [
      {
        "id": "trim",
        "label": "Trim reads",
        "tool": "Galaxy",
        "stage": "preprocess",
        "scope": "bulk",
        "terminal": false,
        "viz_only": false
      },
      {
        "id": "filter",
        "label": "Filter reads",
        "tool": "Galaxy",
        "stage": "preprocess",
        "scope": "bulk",
        "terminal": false,
        "viz_only": false
      },
      {
        "id": "demux",
        "label": "Demultiplex by cell barcode (8 bp)",
        "tool": "custom scripts",
        "stage": "demultiplex",
        "scope": "bulk",
        "terminal": false,
        "viz_only": false
      },
      {
        "id": "map",
        "label": "Map read 2 to genome",
        "tool": "",
        "stage": "align",
        "scope": "per_cell",
        "terminal": false,
        "viz_only": false
      },
      {
        "id": "count",
        "label": "Count reads per gene",
        "tool": "htseq-count",
        "stage": "quantify",
        "scope": "per_cell",
        "terminal": false,
        "viz_only": false
      },
      {
        "id": "tpm",
        "label": "Normalize to tpm",
        "tool": "custom scripts",
        "stage": "normalize",
        "scope": "per_cell",
        "terminal": false,
        "viz_only": false
      },
      {
        "id": "absolute",
        "label": "Convert tpm to absolute copies",
        "tool": "custom scripts",
        "stage": "normalize",
        "scope": "per_cell",
        "terminal": true,
        "viz_only": false
      }
    ],
    "edges": [
      {
        "from": "trim",
        "to": "filter",
        "kind": "sequential"
      },
      {
        "from": "filter",
        "to": "demux",
        "kind": "sequential"
      },
      {
        "from": "demux",
        "to": "map",
        "kind": "sequential"
      },
      {
        "from": "map",
        "to": "count",
        "kind": "sequential"
      },
      {
        "from": "count",
        "to": "tpm",
        "kind": "sequential"
      },
      {
        "from": "tpm",
        "to": "absolute",
        "kind": "sequential"
      }
    ],
    "statistical_model": "Least-squares linear regression of tpm-normalized counts against known ERCC spike-in concentrations to recover absolute mRNA copy numbers; no gene-length normalization is applied because CEL-Seq retains a single 3'-most fragment per transcript."
  }
}