AlienRemover: data and materials

AlienRemover (https://gitlab.pasteur.fr/GIPhy/AlienRemover) is a program to quickly discard alien reads (e.g. exogenous reads, host, cloning vectors) from FASTQ-formatted files. Alien bases are searched using a fast alien k-mer identification algorithm, and the removal criterion is determined by detecting (based on k-mers) a sufficient proportion of successive alien bases within high-throughtput sequencing (HTS) reads. To determine accurate default values for both k-mer length and sucessive alien base proportion, different datasets were build and stored in this repository. ► Athal_PhiX (Arabidopsis thaliana + PhiX reads) This dataset was inferred from the 2x300 Illumina MiSeq reads associated to the SRA accession SRR726611. These HTS reads correspond to the whole genome sequencing of A. thaliana, but also of Escherichia virus PhiX (used for control). Technical adapter and primer oligonucleotides were detected using Minion (see documentation here), leading to the following ones: >R1_TruSeq_Adapter_Index_22 AGATCGGAAGAGCACACGTCTGAACTCCAGTCACcgtacgTAATCTCGTATGCCGTCTTCTGCTTG >R2_TruSeq_Universal_Adapter_rc AGATCGGAAGAGCGTCGTGTAGGGAAAGAGTGTAGATCTCGGTGGTCGCCGTATCATT >poly-A AAAAAAAAAAAAAAA >poly-C CCCCCCCCCCCCCCC Read pairs were clipped and trimmed using AlienTrimmer v2.0. Only read pairs made up by reads of length at least 220 bps were retained. To properly assign each remaining read pair to its taxon, they were aligned using minimap2 against genome assemblies of A. thaliana (GCF_000001735.4) and PhiX174 (GCA_002588795.1). Read pairs that do not align against these two genome assemblies were discarded. This procedure led to 6,378,237 read pairs: 6,343,954 (99.46%) and 34,283 (0.54%) are associated to A. thaliana and PhiX174, respectively. This read dataset corresponds to the two gzipped FASTQ files Athal_PhiX.1.fastq.gz and Athal_PhiX.2.fastq.gz. Each FASTQ block (i.e. 4-line block) associated to A. thaliana and PhiX has its first line starting with @Arabidopsis_thaliana and @PhiX, respectively. ► hCoV19_Hsap (SARS-CoV-2 + Homo sapiens reads) This dataset was inferred from the 300 bps Illumina iSeq single-end reads associated to the SRA accession SRR12782936. These HTS reads correspond to the whole genome sequencing of a SARS-CoV-2 virus isolate, but contains alien reads from its human host. These reads were processed following a similar procedure as previously described (see above). Trimming and clipping were carried out with the following oligonucleotides: >TruSeq_DNA AGATCGGAAGAGCACACGTCTGAACTCCAGTCACcctatggtATCTCGTATGCCGTCTTCTGCTTG >poly-A AAAAAAAAAAAAAAA >poly-C CCCCCCCCCCCCCCC Taxonomic assignation was performed by aligning reads against genome assemblies of H. sapiens (GCF_000001405.28) and SARS-CoV-2 (GCA_009858895.3), leading to a total set of 392,404 single-end reads, made up by 222,945 (56.81%) SARS-CoV-2 and 169,459 (43.19%) and H. sapiens ones. This read dataset corresponds to the gzipped FASTQ file hCoV19_Hsap.fastq.gz. Each FASTQ block associated to SARS-CoV-2 and H. sapiens has its first line starting with @SARS-CoV-2 and @Homo_sapiens, respectively. ► Homo.sapiens.*.kmr (Homo sapiens k-mers) AlienRemover is able to save into a file the different distinct k-mers associated to alien genomes. To quickly detect and remove alien reads within FASTQ files, these saved alien k-mers can next be directly read by AlienRemover, which is useful when dealing with large alien genomes. When analysing the dataset hCoV19_Hsap, the H. sapiens k-mers were therefore computed by AlienRemover and saved. The H. sapiens k-mer sets correspond to the 11 files Homo.sapiens.k$k.kmr, where $k is odd and varies between 11 and 31.AlienRemover（https://gitlab.pasteur.fr/GIPhy/AlienRemover）是一款可快速从FASTQ格式文件中剔除外源reads（如外源序列、宿主序列、克隆载体序列）的工具。其通过高效的外源k-mer识别算法搜索外源碱基，并基于k-mer检测高通量测序（High-throughput sequencing，以下简称HTS）reads中连续外源碱基的占比是否达标，以此作为剔除判定标准。为确定k-mer长度与连续外源碱基占比的合理默认参数，本仓库构建并存储了多组测试数据集。 ► Athal_PhiX（拟南芥+PhiX reads） 本数据集源自SRA登录号SRR726611对应的2×300 bp Illumina MiSeq测序reads。这些HTS reads既包含拟南芥全基因组测序数据，也包含用于质控的大肠杆菌噬菌体PhiX的测序数据。 研究使用Minion工具（详见官方文档）检测到的接头与引物寡核苷酸序列如下： >R1_TruSeq_Adapter_Index_22 AGATCGGAAGAGCACACGTCTGAACTCCAGTCACcgtacgTAATCTCGTATGCCGTCTTCTGCTTG >R2_TruSeq_Universal_Adapter_rc AGATCGGAAGAGCGTCGTGTAGGGAAAGAGTGTAGATCTCGGTGGTCGCCGTATCATT >poly-A AAAAAAAAAAAAAAA >poly-C CCCCCCCCCCCCCCC 随后使用AlienTrimmer v2.0对read对进行接头剪切与序列修剪，仅保留长度≥220 bp的read对。 为将剩余read对准确分配至对应物种，研究使用minimap2将reads比对至拟南芥（GCF_000001735.4）与PhiX174（GCA_002588795.1）的基因组组装序列。未比对至这两个基因组的read对被剔除。最终得到6,378,237条read对，其中6,343,954条（99.46%）属于拟南芥，34,283条（0.54%）属于PhiX174。 本数据集包含两个gzip压缩的FASTQ文件：Athal_PhiX.1.fastq.gz与Athal_PhiX.2.fastq.gz。对应拟南芥与PhiX的FASTQ块（即4行格式的read块）的首行分别以@Arabidopsis_thaliana与@PhiX开头。 ► hCoV19_Hsap（SARS-CoV-2 + 智人reads） 本数据集源自SRA登录号SRR12782936对应的300 bp Illumina iSeq单端测序reads。这些HTS reads为某SARS-CoV-2病毒分离株的全基因组测序数据，但包含来自人类宿主的外源reads。 该数据集的处理流程与前述Athal_PhiX数据集一致。使用以下寡核苷酸序列进行接头剪切与序列修剪： >TruSeq_DNA AGATCGGAAGAGCACACGTCTGAACTCCAGTCACcctatggtATCTCGTATGCCGTCTTCTGCTTG >poly-A AAAAAAAAAAAAAAA >poly-C CCCCCCCCCCCCCCC 通过将reads比对至智人（GCF_000001405.28）与SARS-CoV-2（GCA_009858895.3）的基因组组装序列完成物种分类分配，最终得到392,404条单端reads，其中222,945条（56.81%）属于SARS-CoV-2，169,459条（43.19%）属于智人。 本数据集包含一个gzip压缩的FASTQ文件hCoV19_Hsap.fastq.gz。对应SARS-CoV-2与智人的FASTQ块的首行分别以@SARS-CoV-2与@Homo_sapiens开头。 ► Homo.sapiens.*.kmr（智人k-mer集） AlienRemover支持将外源基因组对应的不同独特k-mer保存至文件中。这些预先保存的外源k-mer可直接被AlienRemover读取，用于快速检测并剔除FASTQ文件中的外源reads，在处理大体积外源基因组数据时尤为实用。在分析hCoV19_Hsap数据集时，AlienRemover即计算并保存了智人的k-mer集。 智人的k-mer集对应11个名为Homo.sapiens.k$k.kmr的文件，其中$k为11至31之间的奇数。