Contig ordering
作用:用 nucmer 把组装 contig 比对到参考组装,推导每个 query contig 的主导参考 contig、比对长度、比例与链向,用于排序与定向。
入口:05.1.nucmer.sh → 05.2.contigOrder.py
输入:Gifu_ref.fa(参考)与 Gifu_contig.fa(组装 contig),文件名硬编码并在工作目录解析。
输出:Gifu_scaffold.delta、Gifu_scaffold.filter.delta、Gifu_scaffold.filter.delta.coords、Gifu_scaffold.delta.coords、一张 mummerplot PNG;mummer_analysis_results/dominant_mapping.csv 与 .../query_contig_on_dominant_ref_colored.pdf(+ 蓝、- 红)。
运行:bash 05.1.nucmer.sh,然后 python 05.2.contigOrder.py
工具:MUMmer nucmer --prefix Gifu_scaffold、delta-filter -i 89 -l 1000 -1、show-coords -THrd、mummerplot --layout --png --large(需 gnuplot);Python 3 + pandas/matplotlib(read_csv(sep='\t', skiprows=5))。
English
Purpose: Align the assembled contigs to a reference with nucmer and derive each query contig’s dominant reference contig, aligned length, ratio and strand, for ordering and orientation.
Entry point: 05.1.nucmer.sh → 05.2.contigOrder.py
Inputs: Gifu_ref.fa (reference) and Gifu_contig.fa (contigs) — hardcoded filenames resolved in the working directory.
Outputs: Gifu_scaffold.delta, Gifu_scaffold.filter.delta, Gifu_scaffold.filter.delta.coords, Gifu_scaffold.delta.coords, a mummerplot PNG; mummer_analysis_results/dominant_mapping.csv and .../query_contig_on_dominant_ref_colored.pdf (+ blue, - red).
Run: bash 05.1.nucmer.sh, then python 05.2.contigOrder.py
Tools: MUMmer nucmer --prefix Gifu_scaffold, delta-filter -i 89 -l 1000 -1, show-coords -THrd, mummerplot --layout --png --large (needs gnuplot); Python 3 with pandas and matplotlib (read_csv(sep='\t', skiprows=5)).
以下为该步骤目录中的实际文件(同步快照,与该目录内容一致)。
06.1.nucmer.sh · Shell · 9 行
nucmer --prefix Gifu_scaffold Gifu_ref.fa Gifu_contig.fadelta-filter -i 89 -l 1000 -1 Gifu_scaffold.delta > Gifu_scaffold.filter.deltashow-coords -THrd Gifu_scaffold.filter.delta > Gifu_scaffold.filter.delta.coords
show-coords -THrd Gifu_scaffold.delta > Gifu_scaffold.delta.coords
mummerplot Gifu_scaffold.filter.delta -R Gifu_ref.fa -Q Gifu_contig.fa --layout --png --large06.2.contigOrder.py · Python · 105 行
import pandas as pdimport matplotlib.pyplot as pltimport osimport seaborn as sns
# 输入输出设置coords_path = "Gifu_scaffold.filter.delta.coords"output_dir = "mummer_analysis_results"os.makedirs(output_dir, exist_ok=True)
# 读取 coords 文件df = pd.read_csv(coords_path, sep='\t', skiprows=5, header=None)df.columns = [ "ref_start", "ref_end", "qry_start", "qry_end", "ref_len", "qry_len", "percent_id", "ref_cov", "qry_cov", "ref_contig", "qry_contig"]
# 方向、长度等辅助列df["strand"] = df.apply(lambda x: "+" if x["qry_start"] <= x["qry_end"] else "-", axis=1)df["alignment_len"] = abs(df["qry_end"] - df["qry_start"]) + 1df["ref_start_clean"] = df[["ref_start", "ref_end"]].min(axis=1)df["ref_end_clean"] = df[["ref_start", "ref_end"]].max(axis=1)
# 汇总每条query对ref的比对长度与方向summary_records = []for qry, subdf in df.groupby("qry_contig"): total_qry_len = subdf["qry_len"].iloc[0] ref_stats = subdf.groupby("ref_contig").agg({ "alignment_len": "sum", "strand": lambda x: list(x) }).reset_index()
for _, row in ref_stats.iterrows(): ref = row["ref_contig"] aligned_len = row["alignment_len"] strands = row["strand"] ratio = aligned_len / total_qry_len plus = strands.count("+") minus = strands.count("-") dominant_strand = "+" if plus >= minus else "-" summary_records.append([qry, ref, aligned_len, ratio, dominant_strand])
summary_df = pd.DataFrame(summary_records, columns=[ "Query_Contig", "Ref_Contig", "Aligned_Length", "Ratio", "Dominant_Strand"])
# 找出每个 query 的 dominant refdominant_mapping = ( summary_df.groupby("Query_Contig") .apply(lambda x: x.loc[x["Aligned_Length"].idxmax(), ["Query_Contig", "Ref_Contig", "Aligned_Length", "Dominant_Strand"]]) .reset_index(drop=True))dominant_mapping["strand"] = dominant_mapping["Dominant_Strand"]dominant_mapping.to_csv(os.path.join(output_dir, "dominant_mapping.csv"), index=False)
# 按 dominant ref 分组,绘制每个 query 的合并线段plot_data = []for _, row in dominant_mapping.iterrows(): qry = row["Query_Contig"] ref = row["Ref_Contig"] strand = row["strand"]
sub = df[(df["qry_contig"] == qry) & (df["ref_contig"] == ref)] if sub.empty: continue start = sub["ref_start_clean"].min() end = sub["ref_end_clean"].max() plot_data.append([ref, qry, start, end, strand])
plot_df = pd.DataFrame(plot_data, columns=["ref_contig", "qry_contig", "start", "end", "strand"])
# 为每个 ref-query 分配独立的 y 坐标plot_df["ref_qry"] = plot_df["ref_contig"] + "|" + plot_df["qry_contig"]y_map = {}y_val = 0for ref in df["ref_contig"].unique(): ref_df = dominant_mapping[dominant_mapping["Ref_Contig"] == ref] for qry in ref_df["Query_Contig"]: key = f"{ref}|{qry}" y_map[key] = y_val y_val += 1plot_df["y"] = plot_df["ref_qry"].map(y_map)
# ytick 按照 ref 显示yticks = plot_df[["y", "ref_contig"]].drop_duplicates().sort_values("y")
# 画图fig, ax = plt.subplots(figsize=(14, 0.5 * len(y_map) + 2))for _, row in plot_df.iterrows(): color = "tab:blue" if row["strand"] == "+" else "tab:red" ax.hlines(y=row["y"], xmin=row["start"], xmax=row["end"], color=color, linewidth=2) ax.text((row["start"] + row["end"]) / 2, row["y"] + 0.3, row["qry_contig"], fontsize=6, ha="center", rotation=45)
ax.set_yticks(yticks["y"])ax.set_yticklabels(yticks["ref_contig"])ax.set_xlabel("Reference Genome Position")ax.set_ylabel("Reference Contigs")ax.set_title("Query Contig Mapping on Dominant Reference Contigs")plt.tight_layout()plt.savefig(os.path.join(output_dir, "query_contig_on_dominant_ref_colored.pdf"))plt.close()
print("✔ dominant_mapping.csv 已保存")print("✔ 主比对关系图 query_contig_on_dominant_ref_colored.pdf 已生成")